The Cholinergic Brain in Parkinson's Disease

Abstract

The central cholinergic system includes the basal forebrain nuclei, mainly projecting to the cortex, the mesopontine tegmental nuclei, mainly projecting to the thalamus and subcortical structures, and other groups of projecting neurons and interneurons. This system regulates many functions of human behavior such as cognition, locomotion, and sleep. In Parkinson's disease (PD), disruption of central cholinergic transmission has been associated with cognitive decline, gait problems, freezing of gait (FOG), falls, REM sleep behavior disorder (RBD), neuropsychiatric manifestations, and olfactory dysfunction. Neuropathological and neuroimaging evidence suggests that basal forebrain pathology occurs simultaneously with nigrostriatal denervation, whereas pathology in the pontine nuclei may occur before the onset of motor symptoms. These studies have also detailed the clinical implications of cholinergic dysfunction in PD. Degeneration of basal forebrain nuclei and consequential cortical cholinergic denervation are associated with and may predict the subsequent development of cognitive decline and neuropsychiatric symptoms. Gait problems, FOG, and falls are associated with a complex dysfunction of both pontine and basal forebrain nuclei. Olfactory impairment is associated with cholinergic denervation of the limbic archicortex, specifically hippocampus and amygdala. Available evidence suggests that cholinergic dysfunction, alongside failure of the dopaminergic and other neurotransmitters systems, contributes to the generation of a specific set of clinical manifestations. Therefore, a “cholinergic phenotype” can be identified in people presenting with cognitive decline, falls, and RBD. In this review, we will summarize the organization of the central cholinergic system and the clinical correlates of cholinergic dysfunction in PD.

The cholinergic neurons of the basal forebrain and the mesopontine tegmental area are the origin of most cholinergic projections in the brain, having extensive connections with the brainstem, striatum, thalamus, hypothalamus, and cortex.¹ Basic neurophysiological evidence, gathered since the mid‐20th century, suggests that the central cholinergic system is involved in regulating many functions of human behavior, such as reward, attention, memory, fear and stress responses, thermoregulation, food intake, and sleep.² In one of the first neuropathological descriptions of Parkinson's disease (PD), Friedrich Lewy reported the presence of typical intraneuronal inclusions in the nucleus basalis of Meynert, at a time when it was not known to contain cholinergic neurons.³ Since the 1970s, many other neuropathological studies have identified Lewy pathology in brain cholinergic nuclei, establishing an association with the clinical manifestations of PD, in particular cognitive decline and locomotor problems. In recent years, in vivo imaging studies of the central cholinergic system, especially positron emission tomography (PET) and magnetic resonance imaging (MRI), have greatly contributed to our understanding of the clinical correlates of cholinergic dysfunction in PD. Signs and symptoms related to cholinergic dysfunction include cognitive decline, gait problems, freezing of gait and falls, REM sleep behavior disorder (RBD), olfactory dysfunction, and neuropsychiatric manifestations. It has been reported that a PD phenotype with RBD, falls and cognitive decline predicts the presence of in vivo cholinergic denervation with reasonable accuracy,⁴ suggesting the existence of a “cholinergic phenotype” in PD. However, many important gaps in our knowledge still remain, such as the lack of known risk factors raising susceptibility to cholinergic dysfunction and its related clinical manifestations.

In this narrative article, we will briefly summarize the anatomic organization of the central cholinergic system and its neuronal markers; then, the clinical correlates of cholinergic dysfunction will be reviewed.

Search Strategy

Pubmed/MEDLINE search for “cholinergic” and “Parkinson” in title or abstract yielded 4282 results from 1948 to 2021. Only articles written in English were examined. For this review, we included studies conducted in humans and investigating the clinical pathophysiology of cholinergic dysfunction in PD, that is, the clinical manifestations associated with cholinergic neurodegeneration. Studies investigating the clinical correlates of cholinergic dysfunction through neuropathological, neuroimaging, and neurophysiological techniques were included. Additional articles with these characteristics were extracted from the references of the included articles. Neuroimaging studies investigating the clinical correlates of cholinergic dysfunction in PD have been approached systematically and the main findings have been summarized in this article's tables.

Neuroanatomical Organization of the Central Cholinergic System

The two main populations of cholinergic neurons in the brain include distal projecting neurons and local interneurons. Cell bodies of most distally projecting cholinergic neurons in the brain originate in the basal forebrain complex, in the pedunculopontine nucleus (PPN) and laterodorsal tegmental nucleus (LDTN) located in the mesopontine tegmental area, and in the medial habenula. Cholinergic interneurons are found in the striatum, nucleus accumbens, and neocortex.², ⁵, ⁶, ⁷ Most investigations of projecting cholinergic nuclei related to human disease, especially Alzheimer's disease (AD) and PD, have focused on the basal forebrain complex, the PPN and the LDTN. The basal forebrain complex includes 4 groups of cholinergic neurons, projecting ipsilateral unmyelinated or thinly myelinated axons: the medial septal nucleus and vertical limb nucleus of the diagonal band (Ch1/Ch2 groups) send axons to the hippocampus; the horizontal limb nucleus of the diagonal band (Ch3 group) provides innervation of the olfactory bulb; and the nucleus basalis of Meynert (Ch4 group) projects diffusely, yet in a discrete and organized manner, to the cerebral cortex and amygdala.⁸, ⁹, ¹⁰, ¹¹ Some nomenclature ambiguity is present in the literature and it is worth stating that we will refer to the cholinergic basal forebrain complex to include all the aforementioned discrete cholinergic nuclei. Moreover, the nucleus basalis of Meynert is sometimes referred to as "the nucleus basalis of Meynert neurons in the substantia innominata", because this nucleus is a scattered group of neurons including magnocellular neurons, 90% of which are cholinergic and lie in the substantia innominata.¹² The PPN and the LDTN are found in the mesopontine tegmental area and contain most, but not all, cholinergic neurons in this region. They provide cholinergic input to the substantia nigra pars compacta (SNpc) and ventral tegmental area (VTA), thalamus, hypothalamus, and, to a lesser extent, to the cerebral cortex⁷, ¹³, ¹⁴, ¹⁵, ¹⁶, ¹⁷ (Fig. 1). A detailed neuroanatomic description of the cholinergic system in the brain of mammals has been reviewed by Mesulam elsewhere.¹⁸ Preclinical studies have highlighted how cholinergic projections, through their extensive connections, modulate many physiological functions: the mesolimbic connections influence response to reward; cortical connections to the prefrontal cortex mediate decision making, planning, and ascribing salience; projections to the hippocampus and amygdala influence attention, memory, fear, and stress responses; whereas hypothalamic connections influence homeostatic responses such as thermoregulation, food intake, and sleep.² Furthermore, a role of the cholinergic system in facilitating synaptic plasticity and neuronal development has been proposed. Cholinergic projections have a modulatory, rather than overtly excitatory or inhibitory, effect on other neuronal systems, and their influence is mediated by complex interactions between pre‐ and post‐synaptic nicotinic and muscarinic receptors. An extensive summary of the neurophysiological processes regulated by the cholinergic system has been published by Picciotto et al.²

FIG. 1.

Cholinergic systems of the brain. Two major pathways project widely to different brain areas: basal‐forebrain cholinergic neurons (red), and pedunculopontine–lateral dorsal tegmental neurons (blue). Other cholinergic neurons include striatal interneurons (orange), cranial‐nerve nuclei (green circles), vestibular nuclei (purple); and spinal cord preganglionic and motor neurons (yellow) (reused with permission from Perry et al¹⁷). Abbreviations: nb: nucleus basalis of Meynert: ms: medial septal nucleus.

Neuronal Cholinergic Markers

Neuronal cholinergic markers have been used in pathological and PET imaging studies to label cholinergic neurons and are reported in Table 1. Knowledge of the localization of these markers will help the reader to understand the methodological aspects of the studies investigating the cholinergic system. Of note, the localization of acetylcholine (ACh) receptors is heterogeneous and complex throughout the central nervous system and a full description is beyond the scope of this review. The subject is reviewed in Nathanson¹⁹ for muscarinic receptors and in Dani and Bertrand²⁰ for nicotinic receptors.

TABLE 1.

Cholinergic receptors and enzyme markers

Cholinergic marker	Description	Marker significance	PET imaging ligands cited in the text
ChAT	Intracellular enzyme responsible for the synthesis of ACh from acetylCoA and choline	Cholinergic axon terminal marker	–
VAChT	Specific ACh transporter that allows the accumulation of ACh into pre‐synaptic vesicles	Cholinergic axon terminal marker	[18F]FEOBV
AChE	Extracellular enzyme present at the cholinergic synapse, responsible for the hydrolytic degradation of ACh in acetate + choline	Cholinergic synaptic marker	[11C]PMP [11C]MP4A [11C]donepezil
Muscarinic receptors (M1—M5)	M1: mostly post‐synaptic M2: mostly pre‐synaptic Different subtypes and localizations in different CNS sites	State of pre‐ and post‐synaptic muscarinic receptors	[11C]NMPB (M1—M2 non‐selective ligand)
Nicotinic receptors (αxβy)	Ion channels receptors on pre‐ (and to a less extent post‐) synaptic neurons. α4β2 is the most common isoform in the CNS	State of pre‐ (and post‐) synaptic nicotinic receptors. Located on acetylcholine, dopamine, serotonin noradrenaline, GABA, and glutamate axon terminals	2‐[18F]FA‐85380 (α4β2 nAChR‐specific) [18F]flubatine

Binding to these markers is used to label cholinergic neurons in neuropathological and PET studies.

Abbreviations: [¹⁸F]FEOBV, fluorine‐18‐labeled fluoroethoxybenzovesamicol; [¹¹C]PMP, carbon‐11‐labeled methyl‐4‐piperidinyl propionate; [¹¹C]MP4A, N‐[¹¹C]methylpiperidin‐4‐yl acetate; [¹¹C]NMPB, carbon‐11‐labeled N‐methyl‐4‐piperidyl benzilate; 2‐[¹⁸F]FA‐85380, 2‐[¹⁸F]fluoro‐3‐(2[S]‐2‐azetidinylmethoxy)‐pyridine; AChE, acetylcholine esterase; ChAT, choline acetyl transferase; GABA, gamma‐aminobutyric acid; VAChT, vesicular acetylcholine transporter.

Clinical Implications of Cholinergic Dysfunction in PD

We will now review the currently available evidence relating to PD manifestations and cholinergic dysfunction. Specifically, the following topics will be reviewed: cognitive impairment, gait problems and falls, sleep problems, neuropsychiatric symptoms, olfactory dysfunction, and peripheral cholinergic parasympathetic system dysfunction. Finally, a brief section will be dedicated to current pharmacological approaches to improve cholinergic dysfunction. It must be acknowledged that all these symptoms have a complex pathophysiology; in this review, we will focus on the cholinergic contributions, referring to other neurotransmitters non‐systematically.

Investigating Cholinergic Dysfunction in the Clinical Setting

Clinical correlates of cholinergic dysfunction have been studied with a variety of approaches. The first approach to the study of the cholinergic system in PD involved post‐mortem pathological examinations. More recently, in vivo neuroimaging and neurophysiological techniques have emerged. Neuroimaging approaches reviewed here include PET and single photon emission computed tomography (SPECT) studies using cholinergic ligands, and MRI studies using sequences that address the integrity of structure and connections of the cholinergic nuclei. The cortical cholinergic system may also be evaluated indirectly through clinical neurophysiology techniques. Short‐latency afferent inhibition (SAI) couples peripheral nerve stimulation with transcranial magnetic stimulation (TMS) of the contralateral motor cortex, giving a proxy measure of cholinergic excitability of the cerebral cortex.²¹, ²² Cholinergic dysfunction leads to a reduction of SAI (ie, a greater reduction of the recorded motor evoked potential when the SAI protocol is applied). This is also seen after intravenous injections of scopolamine, a muscarinic receptor antagonist, and it has been shown to be present in people with AD and PD dementia. Conversely, it may be reversed with pharmacological blockade of acetylcholine esterase. The SAI technique is extensively reviewed by Turco et al²³. Moreover, some quantitative electroencephalographic (EEG) metrics, such as impaired suppression of α rhythm on eye opening, have shown potential as markers of cholinergic dysfunction.²⁴, ²⁵, ²⁶ Some of these alterations also showed reversibility after anticholinesterase inhibitor treatment.²⁷, ²⁸

Cognitive Impairment

Cognitive impairment is frequent in PD and it is estimated that 75% of people who survive longer than 10 years will develop dementia.²⁹ The cholinergic nucleus basalis of Meynert was one of the sites where Lewy bodies were first reported.³ In Braak's pathological staging of PD, basal forebrain pathology occurs simultaneously with nigral pathology.³⁰ Early reports of an association between cholinergic dysfunction and cognitive impairment in PD appeared ~60 years ago, when the mainstay of antiparkinsonian treatment was benzhexol and other muscarinic antagonists before the advent of levodopa, and their cognitive side effects were well recognized.

Neuropathological studies reported neuronal loss and Lewy bodies in the nucleus basalis of Meynert of people with PD.³¹, ³² Other post‐mortem investigations estimated the loss of cholinergic neurons in the nucleus basalis of Meynert to be between 30% and 70%.³³, ³⁴, ³⁵ Furthermore, a severe loss of cholinergic basal forebrain neurons was reported to differentiate demented from non‐demented people with PD³⁶ and a decrease in cortical choline acetyl‐transferase (ChAT), the enzyme found in cholinergic terminals responsible for acetylation of choline, has also been associated with cognitive decline.³¹, ³⁷ More recently, a study from Hall and colleagues³⁸ delved into the details of cholinergic dysfunction and its association with PD dementia. Sixteen brains of people with PD, 8 with and 8 without dementia, and 8 controls were analyzed. Only people without evidence of concomitant other pathologies associated with dementia, such as AD, were included. The Ch1 and Ch2 cholinergic nuclei of the basal forebrain, projecting to the hippocampus, did not show significant cell loss in any participant; however, their terminals were significantly reduced only in demented people, who also had a higher burden of Lewy pathology in the basal forebrain. Furthermore, only people with PD dementia showed significant loss of cholinergic neurons in the nucleus basalis of Meynert (Ch4). Therefore, the authors hypothesized that α‐synuclein deposition in the basal forebrain nuclei, even in the absence of significant neuronal loss, may drive terminal cholinergic dysfunction and, together with widespread Lewy body pathology, the development of dementia.³⁸

Molecular neuroimaging of the cholinergic system has greatly contributed to our understanding of the clinical correlates of cholinergic dysfunction in PD cognitive impairment. One of the first reports was a study investigating the status of the muscarinic receptors (mAChRs) in people with PD using [¹¹C]NMPB‐PET.³⁹ The authors found an increase in mAChR availability in the frontal cortex in PD compared to controls and hypothesized that this resulted from loss of afferent cholinergic innervation. Since then, a series of PET investigations conducted with different radioligands has detailed aspects of the relationship between cholinergic dysfunction and cognitive impairment in vivo. Significant reductions in the cortical presynaptic cholinergic markers acetylcholine esterase (AChE) and vesicular acetylcholine transporter (VAChT) have been found in ~30% of non‐demented, mild‐to‐moderate stage people with PD compared to healthy controls.⁴⁰, ⁴¹ This reduction in signal was quantified as ~10% in cognitively unimpaired PD and 20%–30% in PD dementia (PDD).⁴², ⁴³, ⁴⁴, ⁴⁵, ⁴⁶ Reductions have been reported to target parieto‐occipital regions in non‐demented cases and spread to frontal and temporal cortex in more cognitively affected cases.⁴², ⁴⁶, ⁴⁷ Loss of cortical cholinergic terminals is associated with impaired cognitive performance in non‐demented people with PD⁴⁸; in one study caudate dopaminergic terminal denervation was also an independent predictor of cognitive decline and showed a statistically significant interaction effect with cortical cholinergic denervation. The pathophysiological relationship between cortical cholinergic denervation, caudate dopaminergic dysfunction, and cognitive decline remains to be fully elucidated. Loss of cortical and subcortical cholinergic denervation has not only been associated with impaired visual perception, but also deficits in attention and executive dysfunction.⁴⁴, ⁴⁹ New tracers with a higher signal‐to‐noise ratio will likely allow a better recognition of the relationship between discrete patterns of cholinergic dysfunction and specific cognitive problems.⁵⁰ A summary of the main studies investigating cholinergic dysfunction through molecular neuroimaging may be found in Table 2.

TABLE 2.

Summary of PET studies cited in the text relevant to cholinergic dysfunction and cognitive impairment

Author, year	Imaging technique	Number and types of pts	Findings
Asahina et al, 199539	[11C]NMPB‐PET (M1‐M2 receptors marker)	8 non‐demented PD pts; 8 HCs	Increased binding in the frontal cortex compared with other cerebral cortical areas, 20% higher than controls. No correlation between regional binding and clinical characteristics.
Kuhl et al, 199642	[123I]IBVM‐SPECT (acetylcholine vesicular transporter marker) [18F]FDG‐PET	9 PD, 6 PDD, 22 AD, 36 HCs	HCs: no age‐dependent terminal density; neocortical decline 3.7%/decade. PD: reduced cholinergic terminal density in parietal and occipital cortices. PDD: extensive cortical binding decreases.
Bohnen et al, 200343	[11C]PMP‐PET (AChE marker)	11 PD, 14 PDD, 12 AD, 10 HCs	PD: cortical AChe binding reduction 12.9%. PDD: cortical AChe binding reduction 20.0%. AD: cortical AChe binding reduction 9.1% (with selective involvement of lateral temporal cortex).
Hilker et al, 200545	[11C]MP4A‐PET (AChE marker) [18F]FDOPA‐PET (dopaminergic terminals marker)	17 PD, 10 PDD, 31 HCs	PD: global cortical AChE binding reduction 10.7%. PDD: global cortical AChE binding reduction 29.7%; significant reductions in parietal cortex.
Shimada et al, 200946	[11C]MP4A‐PET (AChE marker)	9 early PD, 9 advanced PD, 10 PDD, 11 DLB, 26 HCs	Early and advanced PD: similar profile of global cortical AChE binding reduction (11.8%), reductions greater in medial occipital cortex. PDD/DLB: similar profile of global cortical AChE binding reduction (PDD: 23.4%; DLB: 27.1%), especially in posterior cortical regions. Significant correlation between MMSE scores and mean cortical binding in the whole cohort.
Bohnen et al, 201240	[11C]PMP‐PET (AChE marker) [11C]DTBZ‐PET (dopamine vesicular transporter marker)	101 non‐demented PD, 29 HCs	Cortical AChE activity below 5th percentile of normal values in 31% of PD pts. Thalamic AChE activity below 5th percentile of normal values in 18% of PD pts. Normal cortical and thalamic binding values 65% of pts.
Bohnen et al, 201548	[11C]PMP‐PET (AChE marker) [11C]DTBZ‐PET (dopamine vesicular transporter marker)	143 non‐demented PD patients classified according to cognitive performances relative to normative values	Increase in frequency of cortical cholinergic denervation in categories of increasing cognitive impairment. Increase in frequency of caudate dopaminergic denervation in categories of increasing cognitive impairment. Significant independent and interactive effects of cortical cholinergic denervation and caudate nucleus dopaminergic in predicting cognitive impairment category.

Abbreviations: [¹¹C]DTBZ‐PET, [¹¹C]dihydrotetrabenazine; [¹⁸F]FEOBV, [¹⁸F]fluoroethoxybenzovesamicol; [¹²³I]IBVM, [¹²³I]iodobenzovesamicol; [¹¹C]MP4A, N‐[¹¹C]‐methyl‐4‐piperidyl actetate; [¹¹C]PMP, N‐[¹¹C]‐methyl‐4‐piperidyl propionate; [¹¹C]NMPB, N‐[¹¹C]‐methyl‐4‐piperidyl benzilate; AChE, acetylcholine esterase; AD, Alzheimer's disease; DLB, dementia with Lewy bodies; FDG, fluorodeoxyglucose; HCs, healthy controls; MMSE, Mini Mental State Examination; PD, Parkinson's disease; PDD, Parkinson's disease dementia; pts: patients; PET, positron emission tomography; WCST, Wisconsin Card Sorting Test.

MRI can also detect pathologic changes of the basal forebrain complex. Atrophy of the substantia innominata region has been shown in PD⁵¹, ⁵² with a greater involvement in PDD, and worse neurocognitive deficits have been associated with greater degrees of atrophy.⁵³ Since more precise analyses of the basal forebrain nuclei have become available, it has been shown that atrophy of the nucleus basalis of Meynert (Ch4) is associated with a higher risk of cognitive decline during follow up54, 55, 56 whereas PPN atrophy is associated with gait disorders.⁵⁷ Advanced techniques such as diffusion tensor imaging examining structural connectivity and free water pools⁵⁸ and functional MRI can identify definite local structural and both structural and functional network alterations related to specific neurocognitive alterations⁵⁹; for example, loss of Ch1‐2 projections to hippocampus has been associated with worse scores on memory tests, whereas changes in Ch3‐Ch4 projections to the neocortex have been associated with worse scores on Mini Mental State Examination (MMSE) and executive functions tests. Finally, recent studies have highlighted how decreased basal forebrain volume in PD is associated with quantitative EEG metrics commonly associated with dementia, such as reduced α rhythm reactivity, showing that EEG abnormalities could be closely related to cortical cholinergic dysfunction.²⁵, ²⁶ A summary of the main studies investigating cholinergic dysfunction through MRI may be found in Table 3.

TABLE 3.

Summary of MRI studies cited in the text relevant to cholinergic dysfunction and cognitive impairment

Author, year	Imaging technique	No. and types of patients	Findings
Oikawa et al, 200452	1.5 T MRI (T2 FSE)	44 PD (4 pts with MMSE <23), 20 HCs	SI thickness in PD significantly reduced compared to HCs, especially in pts with dementia. Significant positive correlation between SI thickness and MMSE.
Choi et al, 201253	3 T MRI (3D T1 TFE)	24 PD, 35 PD‐MCI, 29 PDD, 28 HCs	SI volume significantly decreased in all pts compared to controls, and lower in PDD compared to PD and PD‐MCI. SI volume significantly associated with cognitive status and performance in cognitive subdomains, especially attention and object naming.
Ray et al, 201854	3 T MRI (3D T1 MP‐RAGE)	168 de novo PD, 76 HCs MRI baseline analysis, 5‐yr longitudinal clinical analysis	Ch4p volume reduced in PD pts with MCI. Reduced Ch4 volume independently increases the risk of MCI or PDD in the subsequent 5 yr. No significant differences in cognitive performances at baseline between patients with normal and reduced Ch4 volume; pts with reduced Ch4 volume had faster decline in cognitive performances.
Gargouri et al, 201959	3 T MRI (3D T1, 3D T2, diffusion, resting state fMRI)	52 non‐demented PD (20 MCI), 25 HCs	Abnormal structural indices (mean, axial and radial diffusivity) in Ch1‐2‐3‐4. Reduced structural connectivity between Ch3‐4 and associative prefrontal, occipital, peri‐insular cortices. Reduced functional connectivity between Ch1‐2 and hippocampi and between Ch3‐4 and frontal areas and thalamus. Associations between structural and functional connectivity indices and specific cognitive performances.
Barrett et al, 201955	3 T MRI (3D T1 MP‐RAGE)	228 de novo PD, 125 advanced PD, 101 HCs	De novo PD had higher Ch123‐4 gray matter density compared to advanced PD. Ch123‐4 gray matter density associated with cognitive performances. In advanced PD, Ch123‐4 nuclei gray matter density was associated with MoCA and cognitive subdomains performances. Ch4 density associated with attention and visuospatial function in both groups.
Gang et al, 202056	1.5 T MRI (3D structural) [18F]FDG‐PET	56 PD, 13 HCs Baseline and 3‐yr assessments	At baseline 20 PD pts with nBM atrophy. Pts with nBM atrophy showed more severe cognitive, anxiety and apathy symptoms both at baseline and 3‐yr follow up. They also showed significantly reduced FDG uptake in parietal and occipital cortices at baseline and follow up.

Abbreviations: AD, Alzheimer's disease; Ch123‐4, cholinergic groups 1–2, 3, and 4 of the basal forebrain; Ch4p, posterior part of the Ch4 cholinergic group of the basal forebrain; DLB, dementia with Lewy bodies; FDG, fluorodeoxyglucose; fMRI, functional MRI; FSE, fast spin; echo; HCs, healthy controls; MMSE, Mini Mental State Examination; MP‐RAGE, magnetization‐prepared rapid acquisition gradient echo; MRI, magnetic resonance imaging; nBM, nucleus basalis of Meynert; PD, Parkinson's disease; PDD, Parkinson's disease dementia; pts, patients; SI, substantia innominata; T, Tesla; TFE, turbo field echo.

Overall, neuropathological and neuroimaging studies have provided strong evidence linking cholinergic dysfunction and cognitive decline in PD. However, some relevant questions remain unanswered; because not all people with PD develop overt dementia, which factors provide increased risk of cholinergic dysfunction progression and the development of cognitive decline? Because cholinergic dysfunction is present in some, but not all people with PD, is it associated with more widespread α‐synuclein pathology, and is this in turn associated with an increased risk of developing dementia?

Gait Problems and Falls

Compared to healthy subjects, PD gait is characterized by reductions and variability in step length and loss of automatic movement,⁶⁰ features that have been linked to increased risk of falls.⁶¹ As a result, people with PD have to pay increased attention when walking⁶² and gait becomes a challenging exercise, especially if performing simultaneous motor or cognitive tasks.⁶³, ⁶⁴ The added presence of executive and attentional deficits related to cortical cholinergic terminal dysfunction act to exacerbate gait problems.⁶⁵

Early neuropathological evidence suggested that PPN degeneration plays a role in gait and balance dysfunction in PD and its atypical variants.⁶⁶ The importance of the PPN cholinergic neurons in locomotion has been extensively explored in preclinical studies, recently reviewed in French and Muthusamy.⁶⁷

Recent evidence from in vivo clinical and neuroimaging studies has contributed to the understanding of the complex processes that lead to gait problems and falls. Using [¹¹C]PMP PET, Bohnen and colleagues have measured thalamic AChE activity, which mainly stems from PPN projections, and found it significantly reduced in those with a history of falls compared to “non‐fallers”, regardless of the degree of nigrostriatal denervation.⁶⁸ In a follow up investigation, the authors used [¹⁸F]FEOBV PET, a marker of VAChT, to confirm and extend those findings: cholinergic denervation of the right posterior thalamus, a relay in the visual network, was especially associated with history of falls, suggesting that impaired processing of visual information contributes to gait problems.⁶⁹ Thalamic cholinergic deficits are also associated with impaired postural reflexes, adding to the pathophysiological link between PPN‐thalamic cholinergic denervation and falls.⁷⁰

A recent MRI investigation involving the Parkinson Progressive Makers Initiative (PPMI) cohort showed that baseline microstructural changes of the PPN, namely increased axial diffusivity detected with diffusion tensor imaging (DTI), and reduced caudate dopaminergic availability were independently associated with a higher burden of postural instability/gait difficulty symptoms in the subsequent 5 years of follow‐up, whereas structural changes of the nucleus basalis of Meynert (nBM) were not associated.⁵⁷ However, because of the interplay between cognitive and gait function, loss of cortical cholinergic innervation stemming from basal forebrain cholinergic projections has been indirectly implicated in gait dysfunction. A [¹¹C]PMP PET study has shown that significantly decreased cortical AChE activity is associated with decreased gait speed, whereas people with PD with intact cortical cholinergic innervation have similar gait speed to controls. This effect was independent of thalamic AChE activity and of the degree of nigrostriatal denervation.⁴¹ Similarly, two recent MRI studies in PD have shown that loss of nucleus basalis of Meynert volume and gray matter density are associated with typical changes in gait parameters (increased step time variability, shortened swing time, and step length) and reduced gait speed.⁷¹, ⁷²

The occurrence of freezing of gait (FOG) has also been linked to cholinergic dysfunction, although other factors such as nigrostriatal dopaminergic dysfunction likely play a substantial role.⁷³ Indeed, one multitracer PET imaging study with [¹¹C]PMP and [¹¹C]PIB has shown that cortical cholinergic denervation is associated with increased risk of manifesting freezing of gait, especially in the presence of concomitant cortical amyloid deposition; no influence of PPN cholinergic thalamic projections on FOG was shown in this study.⁷⁴ A more recent study noted that cholinergic terminal density was significantly reduced in the striatum, hippocampus, and amygdala of “freezers” compared to “non‐freezers”.⁶⁹ It was suggested that loss of caudate nucleus cholinergic interneurons play a key role in the manifestation of freezing of gait. Other series, however, have found MRI abnormalities of the PPN in people with freezing of gait, including gray matter atrophy and an increased free water pool.⁵⁸, ⁷⁵ Increased functional fMRI activity during motor imagery of gait has been reported, and DTI has shown reduced structural connectivity between the PPN and the cerebellum,⁷⁶ the thalamus and multiple regions of the frontal, and prefrontal cortex.⁷⁷ These studies highlight the complexity of network and neurotransmitter abnormalities associated with freezing of gait, in which the cholinergic system dysfunction is likely to be only one of the factors at play. Detailed summaries of neuroimaging studies investigating gait problems may be found in Table 4.

TABLE 4.

Summary of neuroimaging studies cited in the text relevant to cholinergic dysfunction and gait problems

Author, year	Imaging technique	No. and types of pts	Findings
Bohnen et al, 200968	[11C]PMP‐PET (AChE marker) [11C]DTBZ‐PET (dopamine vesicular transporter marker)	44 non‐demented PD, 15 HCs	Cortical AChE binding reduced in “fallers” (−12.3%) and non‐“fallers” (−6%) compared to controls. Thalamic AChE binding reduced only in “fallers” (−11.8%). No differences in nigrostriatal denervation.
Müller et al, 201370	[11C]PMP‐PET (AChE marker) [11C]DTBZ‐PET (dopamine vesicular transporter marker)	124 PD, 25 HCs	Decreased thalamic cholinergic AChE binding associated with decreased postural control. No significant effects of cortical cholinergic or striatal dopaminergic.
Bohnen et al, 201341	[11C]PMP‐PET (AChE marker) [11C]DTBZ‐PET (dopamine vesicular transporter marker)	125 non‐demented PD	Significantly lower cortical cholinergic AChE binding is associated with reduced gait speed, regardless of nigrostriatal denervation and thalamic AChE binding.
Bohnen et al, 201474	[11C]PMP‐PET (AChE marker) [11C]PIB (β‐amyloid ligand) [11C]DTBZ‐PET (dopamine vesicular transporter marker)	143 predominantly non‐demented PD	FOG was most frequent (41.7%) in pts with cortical cholinergic denervation and β‐amyloid accumulation, intermediate (14.3%) in pts with either one, and lowest (4.8%) in pts without cortical cholinergic denervation and β‐amyloid accumulation. No effect of thalamic cholinergic denervation.
Bohnen et al, 201969	[18F]FEOBV (ACh vesicular transporter marker)	94 non‐demented PD	Reduced thalamic cholinergic terminal density in “fallers” compared to “non‐fallers”, especially in right visual thalamus. Reduced cholinergic terminal density in striatum and limbic archicortex of pts with FOG compared to pts without.
Craig et al, 202057	MRI (T1‐ and diffusion‐weighted) 123Ioflupane SPECT	147 de novo PD, 65 HCs Baseline MRI and 5‐yr longitudinal clinical assessments.	PPN microstructural alterations (increased axial diffusivity) are associated with development of PIGD symptoms after 5 yr. No associations for nbM. Reduced caudate DAT binding also was, to a lesser extent, predictive of PIGD symptoms.
Wilson et al, 202171	3 T MRI (T1 MP‐RAGE)	99 PD, 47 HCs 36 months longitudinal follow up	No differences in cholinergic basal forebrain volumes between PD and controls. Ch4 and Ch4p baseline volumes predicted worsening in typically abnormal PD gait parameters after 3 yr.
Dalrymple et al, 202172	3 T MRI (T1 MP‐RAGE)	66 PD (pre‐surgical pts)	Reduced Ch4 gray matter density was associated with slower gait speed and greater worsening during dual tasks.

Abbreviations: [¹¹C]DTBZ‐PET, [¹¹C]dihydrotetrabenazine; [¹⁸F]FEOBV, [¹⁸F]fluoroethoxybenzovesamicol; [¹¹C]Pittsburgh, compound‐B; [¹¹C]PMP, N‐[¹¹C]‐methyl‐4‐piperidyl propionate; AChE, acetylcholine esterase; Ch4p, posterior portion of Ch4 cholinergic group of the basal forebrain; DAT, dopamine transporter; FOG, freezing of gait; FDG, fluorodeoxyglucose; HCs, healthy controls; MP‐RAGE, magnetization‐prepared rapid acquisition gradient echo; MRI, magnetic resonance imaging; nbM, nucleus basalis of Meynert; PIGD, postural instability/gait difficulty; PD, Parkinson's disease; PPN, pedunculopontine nucleus; pts, patients; T, Tesla.

Overall, the evidence presented in this review strongly suggests that the dysfunction and degeneration of basal forebrain and pontine cholinergic nuclei in PD create a complex and inter‐related state of cognitive, postural, and locomotion abnormalities. It must be noted that these are also influenced by dysfunction of other networks. To better clarify the pathophysiology of these manifestations, future research should aim to untangle the relationship between the dysfunction of the different cholinergic nuclei, cognitive deficits, and changes in gait parameters, while taking into account the degeneration of other neurotransmitter systems.

Neuropsychiatric Symptoms

Neuropsychiatric symptoms in PD include depression, anxiety, visual hallucinations and psychosis, fatigue, and apathy.⁷⁸ Depression, visual hallucinations, and psychosis are frequent symptoms that have been linked to cholinergic dysfunction, although changes in other neurotransmitter dysfunctions, such as dopamine loss and use of replacement therapy, also play a role.⁷⁹, ⁸⁰ One PET study found an association between reduced cortical AChE activity and the presence of more severe depressive symptoms in a group of 28 demented and non‐demented males with PD.⁸¹ Another study used 2‐[¹⁸F]FA‐85380 PET to investigate availability of nicotinic α₄β₂ receptors in the brain.⁸² The α₄β₂ nicotinic receptors have a broad distribution throughout the CNS and are located mostly on presynaptic cholinergic, dopaminergic, noradrenergic, serotoninergic, glutamatergic and GABAergic terminals, axons and soma, with a modulating role on the release of these neurotransmitters.²⁰ Meyer and colleagues found a reduction of α₄β₂ receptor availability in the anterior cingulate cortex of people with depressive symptoms compared to those without. The severity of depressive symptoms was also associated with α₄β₂ receptor availability in the anterior cingulate cortex, putamen, and midbrain, areas that are part of a cortico‐mesostriatal circuit.⁸² Notably, in both PET studies an interplay between the cholinergic system, depression, and cognitive decline was highlighted, with cognitive decline associated with more severe depressive symptoms. Indeed, it has previously been suggested that depression in PD may be a risk factor for the subsequent development of dementia,⁸³ therefore, cholinergic dysfunction may be a common denominator of both manifestations.

Visual hallucinations and delusions are features of PD psychosis and occur frequently in association with cognitive decline as disease progresses.⁸⁴ These manifestations may have an underlying cholinergic basis, since pharmacological blockade of muscarinic receptors can induce hallucinations.⁸⁵ Recently, a link between cholinergic dysfunction and psychotic symptoms has been suggested. Manganelli et al⁸⁶ have applied short‐latency afferent inhibition (SAI), a neurophysiological technique used as a proxy measure of cholinergic function, to people with PD with and without visual hallucinations. The former had significantly more abnormal SAI and impaired visuospatial and attentional functions, indicating a possible link between cholinergic abnormalities, visual hallucinations, and specific cognitive deficits.⁸⁶ Furthermore, MRI studies have shown gray matter volume reductions in cholinergic areas of the PPN and the substantia innominata in participants with visual hallucinations compared to those without.⁸⁷, ⁸⁸ Reduced gray matter density in the nucleus basalis of Meynert has also been associated with the subsequent development of psychotic manifestations.⁸⁹

Although the number of studies investigating the relationship between neuropsychiatric symptoms and cholinergic dysfunction in PD is limited, a pathophysiological link appears established. As with other symptoms related to cholinergic dysfunction, an interplay between cholinergic dysfunction, neuropsychiatric symptoms, and cognitive impairment has been suggested.

Sleep Problems

Rapid eye movement (REM) sleep behavior disorder (RBD) is a sleep disorder associated with synucleinopathies, occurring in ~30% of early‐ and intermediate‐stage people with PD.⁹⁰ Isolated RBD (iRBD) is a prodromal neurodegenerative condition in which the characteristic failure to lose muscle tone during REM sleep occurs years before the manifestation of a synucleinopathy.⁹¹ RBD results from the dysfunction and degeneration of the network controlling REM sleep atonia, a circuit that includes the PPN‐LDTN cholinergic neurons.⁹² Experimental activation of the PPN‐LDTN cholinergic neurons is capable of inducing, but not maintaining, REM sleep, and the basal forebrain cholinergic neurons may play a role in REM sleep maintenance.⁹³

One [¹¹C]PMP PET study found decreased AChE binding in thalamic, cortical limbic, and neocortical areas in people with PD and RBD compared to those without RBD; no between‐group differences were found in nigrostriatal dopaminergic function or brainstem and striatum serotonin reuptake transporter availability.⁹⁴ These findings highlight a possible pathophysiological link between cholinergic dysfunction and the manifestation of RBD in PD. However, alterations in cholinergic transmission could also be regarded as the expression of a more severe disease phenotype, because people with RBD have been shown to have a more widespread synucleinopathy⁹⁵ and a higher risk of global disease progression.⁹⁶ In this study, no significant differences were present in motor and cognitive scores between the 2 groups, but a follow up analysis is not available. Therefore, it cannot be excluded that RBD and cholinergic dysfunction constituted the initial stage of a more severely progressing phenotype.

Cholinergic dysfunction may also occur in iRBD. A [¹¹C]donepezil PET study showed a mean reduction of 7.65% in neocortical binding, with 3 of 17 participants having AChE binding 2 SDs below the control mean; a voxel‐wise analysis also showed that people with a Montreal Cognitive Assessment (MoCA) score <27 had more significant [¹¹C]donepezil binding reductions in frontal, occipital, and temporal areas compared to those with scores >26.⁹⁷ Interestingly, a subsequent PET study in people with iRBD showed an inverse correlation between cortical uptake of [¹¹C]donepezil and substantia innominata binding of [¹¹C](R)‐PK11195, a translocator protein ligand and marker of microglial activation. This suggests that higher levels of neuroinflammation in the basal forebrain are associated with decreased cortical cholinergic innervation.⁹⁸ [¹⁸F]FEOBV PET is a marker of ACh vesicular transporter availability and provides an in vivo measure of cholinergic terminal density. In a study of 5 people with iRBD and 5 controls the authors reported increased cholinergic terminal function in several subcortical brain structures known to be involved in REM sleep regulation, whereas cortical areas did not show binding abnormalities. The authors interpreted these findings as a possible compensatory mechanism because of initial dysfunction of acetylcholine production.⁹⁹ Indeed, in iRBD, reduced AChE activity and increased cholinergic terminal density could both represent compensatory mechanisms to maintain cholinergic function. In this regard, one ¹¹C‐MP4A PET study investigated AChE activity in LRRK2 PD and LRRK2 asymptomatic mutation carriers. Interestingly, increased activity in some cortical regions and the thalamus was detected in mutation carriers compared to controls, and in LRRK2 PD compared to idiopathic PD.¹⁰⁰ Although the hypercholinergic state might be a compensatory phenomenon of initial cholinergic denervation, it has also been hypothesized that increased tracer retention could be related to increased inflammation.¹⁰¹ Indeed, microglial cells are known to express the α₇ nicotinic receptor and its activation downregulates their proinflammatory responses in vitro.¹⁰² Furthermore, cerebral butyrylcholinesterase (also called pseudocholinesterase, a hydrolase for esters of choline such as acetylcholine, succinylcholine etc.) is also known to modulate neuroinflammation.¹⁰³ Therefore, these findings of increased cholinesterase activity may be closely related to a proinflammatory state.

As several neuropathological and neuroimaging investigations have suggested that cholinergic dysfunction may occur early in PD,³⁰, ⁴⁶, ⁵⁴, ⁵⁵, ¹⁰⁴ these studies provide even further insights into the mechanisms of early cholinergic dysfunction in synucleinopathies and suggest that cholinergic nuclei degeneration may start before the onset of motor symptoms.

Olfactory Dysfunction

Loss of smell is frequent in PD and may also be a prodromal feature.¹⁰⁵, ¹⁰⁶ Although olfactory performance was initially thought to be associated with nigrostriatal denervation,¹⁰⁷ more recent evidence has also highlighted a pathophysiological connection with cholinergic dysfunction. In a [¹¹C]PMP PET study measuring AChE brain activity in 58 people with PD, a significant association was found between PMP binding in limbic archicortex, i.e. hippocampus and amygdala, and olfactory UPSIT scores; conversely no significant associations were found with severity of nigrostriatal denervation or other demographic characteristics used as co‐regressors.¹⁰⁸ More recently, two studies used SAI as a neurophysiological index of cholinergic dysfunction and found an association between abnormal SAI and olfactory impairment.¹⁰⁹, ¹¹⁰ Notably, Versace and colleagues¹¹⁰ used olfactory‐related evoked potentials as a measure of olfactory dysfunction instead of the more commonly used psychophysical testing (eg, sniffing tests); they also found mild cognitive impairment in most subjects with frank olfactory dysfunction. Indeed, the presence of severe olfactory dysfunction has previously been linked to greater cognitive difficulties in processing speed and verbal abilities¹¹¹ and to increased risk of progression to overt dementia.¹¹² Although a clear pathophysiologic mechanism of olfactory impairment in PD has yet to be established, contributions from Lewy body pathology in olfactory structures and cholinergic dysfunction appear crucial in the development and progression of this manifestation. Because olfactory dysfunction appears closely related to the development of cognitive decline, it is possible that the spread of the pathological process involving the basal forebrain and disrupting cholinergic transmission evolves over time and contributes to the progression of both clinical manifestations.

Peripheral Cholinergic Dysfunction

In the peripheral nervous system, acetylcholine is the neurotransmitter of all preganglionic autonomic fibers, of postganglionic parasympathetic terminals, of many intrinsic enteric neurons and of lower motor neurons. The major source of parasympathetic preganglionic fibers of the heart and gastrointestinal system is the dorsal vagal motor nucleus, with the sacral intermediolateral column (sacral parasympathetic nucleus) supplying the distal part of the colon.¹¹³ In PD, impaired autonomic dysfunction may be present in early stages and even in prodromal stages.¹¹⁴ Constipation, impaired gastric emptying, decreased saliva production, and decreased heart rate variability are parasympathetic dysfunction manifestations that can be present in de novo PD.¹¹⁵

Neuropathological studies have shown a 50% neuron loss compared to controls in the dorsal vagal motor nucleus in PD,¹¹⁶ and α‐synuclein is found in the sacral cholinergic preganglionic neurons.¹¹⁷ Conversely, the intrinsic enteric plexuses display only minor pathological alterations.¹¹⁸ Accumulation of phosphorylated α‐synuclein was also shown in the gastrointestinal organs in PD and in prodromal stages.¹¹⁹, ¹²⁰ Recently, imaging with [¹¹C]donepezil PET showed reductions of AChE availability in small intestine, colon, and pancreas in early‐ and intermediate‐stage people with PD compared to controls.¹²¹, ¹²² At earlier stages greater reductions were seen especially in the colon, a finding that has been shown also in participants with iRBD.¹²³ These studies highlighted how reduced AChE, especially of the colon, may be a marker of early and prodromal disease, although diagnostic accuracy in distinguishing controls from people with PD was only 82%.¹²²

Overall, as the involvement of the peripheral autonomic system may precede the onset of motor symptoms, future research should aim at consolidating and expanding knowledge about pathophysiological mechanisms of autonomic dysfunction in the early and prodromal stages of PD.

Pharmacological Treatment of Cholinergic Dysfunction

Although a systematic description on cholinergic medications in PD is beyond the scope of this review, we will briefly review the evidence regarding pharmacological manipulation of the central cholinergic system. Detrimental effects of central muscarinic antagonists on cognitive performances are historically known and exemplified by a recent study in older adults.¹²⁴ The use of anticholinergic medications was associated with significantly decreased total brain volume and temporal lobe cortical thickness, reduced glucose metabolism in the hippocampus and lower scores on neuropsychological testing, especially immediate memory recall and executive functioning.

Based on the evidence summarized in this review, cholinergic transmission potentiation might have the potential to ameliorate many symptoms of PD. Indeed, cholinesterase inhibitors (CIs) are widely used in PD dementia. One recent meta‐analysis found CIs effective in treating cognitive impairment, but could not demonstrate an effect on the risk of falls.¹²⁵ No pronounced motor side effects were found, although the risk of manifesting tremor was increased. It must be noted that only 4 studies were eligible for this analysis.

Evidence regarding CIs for improving gait, balance, and the risk of falls is limited but expanding rapidly. Two randomized, double‐blind trials of 150 and 23 subjects, testing rivastigmine and donepezil, respectively, found significant reductions in the number of falls.¹²⁶, ¹²⁷ Patients enrolled in these trials had a known history of falls. Conversely, donepezil 10 mg/day did not improve static and dynamic balance in a cohort of PD people with no or a limited fall history and intact cognition.¹²⁸ One large phase 3 placebo‐controlled trial with rivastigmine is evaluating potential benefits on falls, gait speed and FOG.¹²⁹ Two recent small randomized crossover trials involving ~20 patients each, used levodopa plus donepezil or placebo to study the effect on gait parameters and prefrontal cortex activity.¹³⁰, ¹³¹ Interestingly, prefrontal cortex activity was investigated through functional near‐infrared spectroscopy, a non‐invasive technique based on head caps that use near‐infrared light to measure hemodynamic changes in cortical cerebral blood flow, giving a neurophysiological measurement similar to that of functional MRI. Findings of these 2 trials suggested that adding donepezil to levodopa improved walking and turning parameters while producing changes in prefrontal cortex activity. These results are conceptually promising although the relationship between such effects has yet to be investigated in detail.

A different approach has been shown to be viable in a recent study that used varenicline, a partial agonist of the α₄β₂ nicotinic receptor, in 28 people with PD.¹³² This medication acts as an agonist in cholinergic denervated regions, and as antagonist in cholinergic intact regions. In people with documented cholinergic deficit, varenicline produced a 60%–70% receptor occupancy, improved some gait parameters, and reduced the differences in gait performance between dual task and no‐dual task conditions. In agreement with the latter finding, performance in attention tests reflecting cholinergic functions also improved. As opposed to the initial hypothesis, varenicline use was associated with slower gait speed and had no effect on postural stability. A small yet statistically significant increase in MDS‐UPDRS III was also detected with varenicline treatment; although it was below the threshold of a clinically important worsening,¹³³ future studies with this agent should further investigate the characteristics of this effect. Overall, in terms of selection of candidates for cholinergic potentiation, these studies seem to suggest that a marker to identify patients with cholinergic dysfunction should be used to achieve an optimal benefit–risk ratio.

As previously discussed, cholinergic neurotransmission dysfunction has been shown both in RBD associated with PD and iRBD. Although current treatment of nocturnal symptoms with melatonin and/or clonazepam usually achieves good outcomes, cholinesterase inhibitors have been trialed in resistant people. In a double‐blind, crossover pilot trial of 12 people with PD and RBD symptoms resistant to at least melatonin 5 mg and clonazepam 2 mg, rivastigmine 4.6 mg significantly reduced the mean frequency of RBD episodes over 3 weeks of treatment.¹³⁴

Finally, the usefulness of cholinesterase inhibitors on other symptoms related to cholinergic dysfunction is not well established. It must be acknowledged that many of these symptoms, such as neuropsychiatric ones, are also importantly influenced by dysfunction of other neurotransmitters and by medication.

Conclusions

The cholinergic neurons of the mesopontine tegmental area and the basal forebrain send projections throughout the brain, regulating many discrete functions of human behavior. A large amount of evidence now suggests that, along with dopaminergic loss, cholinergic dysfunction plays a substantial role in many PD clinical manifestations, such as cognitive impairment, gait problems, freezing of gait, falls, REM sleep behavior disorder, depression, visual hallucination, psychosis, and olfactory impairment. These findings suggest that cholinergic dysfunction in PD could be contributing to a specific phenotype, in which such manifestations are more prominent. As a stark example of this, one study found that the combination of RBD symptoms and a history of falls was able to predict combined thalamic and cortical cholinergic deficits with an accuracy of 81%, and the combination of decreased walking speed and a MoCA score ≤24 was able to predict cortical‐only cholinergic denervation with a similar accuracy.⁴

However, some critical questions remain unanswered. For example, the risk factors for the development of a “cholinergic phenotype” in PD are currently not known. Furthermore, most of the available knowledge about cholinergic dysfunction of the striatal interneurons, possibly related to severity of motor symptoms such as tremor and levodopa‐induced dyskinesia, comes from animal studies and lacks clinical translation. Finally, how cholinergic degeneration fits in new hypotheses of disease progression is currently unknown. Indeed, whereas Braak's pathological ascending model postulates a simultaneous involvement of dopaminergic and cholinergic neurons,³⁰ a new model based on in vivo neuroimaging hypothesizes the possibility of ascending or descending gradients of pathology.¹³⁵ However, the latter did not include the study of the central cholinergic system, which still lacks a full characterization in terms of in vivo progression.

Therefore, future research should aim to investigate the risk factors for the development of cholinergic dysfunction and validate the link with a putative “cholinergic phenotype”, possibly in early and prodromal people with parkinsonism. Furthermore, interventional symptomatic and disease‐modifying studies should consider evaluating the integrity of the cholinergic system alongside that of the nigrostriatal pathway. Finally, new treatments strategies for improving cholinergic dysfunction seem feasible and should be further investigated.

Author Roles

(1) Research project: A. Conception, B. Organization, C. Execution; (2) Statistical Analysis: A. Design, B. Execution, C. Review and Critique; (3) Manuscript Preparation: A. Writing of the First Draft, B. Review and Critique.

J.P.: 1B, 1C, 3A, 3B

D.J.B.: 1A, 1B, 3B

N.P.: 1A, 1B, 3B

Disclosures

Ethical Compliance Statement

The authors confirm that the approval of an institutional review board and patient consent were not required for this work. We also confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this work is consistent with those guidelines.

Funding Sources and Conflict of Interest

The authors declare that they did not receive specific funding for this work. The authors declare that there are no conflicts of interest relevant to this work.

Financial Disclosures for the Previous 12 Months

J.P. is supported by the European Academy of Neurology Fellowship program. N.P. and D.J.B report no disclosures.