Greater occipital nerve stimulation for cognitive impairment in Parkinson's disease: A two-stage feasibility and tolerability study

Introduction

Parkinson’s disease (PD) is a progressive neurodegenerative condition and a significant source of global neurological disease burden, with incidence rates increasing over the past 20 years¹. While often characterised primarily as a motor syndrome, non-motor symptoms, including cognitive impairment, contribute significantly to the disability arising from PD, with between 60–80% of people with PD experiencing cognitive impairment that impacts their quality of life². Studies have shown that patients may find non-motor symptoms, including cognitive impairment, more disabling and more difficult to address with medication than traditional motor symptoms³.

Mild cognitive impairment (MCI) may be a presenting feature of PD or may develop over the course of the disease. Studies have found that between 25–45% of non-demented people with PD meet Movement Disorder Society criteria for PD-MCI^4,5. Of PD patients who survive 10 years, 46% will develop PD dementia (PDD), a prevalence that rises to 83% in those who survive 20 years⁶. Thus, cognitive impairment in PD represents an important target of research in terms of disease burden, and PD-MCI may represent a pivotal interventional stage for slowing or preventing the progression of impairment to the point of dementia.

Licensed treatments for cognitive impairment in PD are limited. Rivastigmine, an acetylcholinesterase inhibitor, is licensed for symptomatic treatment in PDD but lacks evidence of benefit in the PD-MCI group⁷. Trials of other pharmacological agents, including rasagiline and atomoxetine, have failed to provide evidence of objective benefit in PD-MCI^8,9. A recent Cochrane review concluded that there is an urgent need for more research in non-pharmacological interventions to address cognitive impairment in PD².

The locus coeruleus (LC) is a brainstem nucleus and the primary source of noradrenergic transmission throughout the brain¹⁰. It has roles in arousal, cognition and learning¹¹, and the role of LC degeneration in PD dementia has been long established¹². PD patients with MCI have been demonstrated to have significantly more evidence of LC dysfunction than those without, as measured by neuromelanin-sensitive MRI¹³. LC degeneration in PD patients correlates specifically with cognitive, as opposed to motor, symptoms¹⁴. Neuromodulatory interventions, including transcranial direct current stimulation (tDCS), repetitive transcranial magnetic stimulation (rTMS), and invasive deep brain stimulation (DBS), have been explored in PD-MCI. Trials to date have been heterogeneous in terms of patient selection, stimulation protocol, outcome measurement, and results. Where results have been positive, these have been inconsistent across studies, with some finding modest benefit in single cognitive domains or tests only^15,16. A 2022 review concluded that neuromodulation for cognitive impairment in PD held promise, but that further work in identifying the most effective and well-tolerated targets was needed¹⁷.

As interest in non-pharmacological interventions increases, novel neuromodulatory approaches have been described. Targeting peripheral afferent nerves to modulate central nervous system neurotransmission and plasticity has emerged as a potentially promising approach. Transcutaneous vagus nerve stimulation (VNS) has demonstrated efficacy in enhancing associative memory in older adults¹⁸. In a novel randomized control trial by Luckey et al., greater occipital nerve stimulation (GONS) utilizing tDCS has similarly been shown to enhance performance of older adults in an associative memory task, with benefits persisting for up to 28 days¹⁹. GONS is hypothesized to enhance cognitive performance by causing downstream excitation of the LC, with an associated stimulation in noradrenergic transmission in the pons, hippocampus, and amygdala. Improvement in associative memory following GONS was associated with increased levels of salivary α-amylase levels, suggesting that increased noradrenergic activity via the locus coeruleus mediated the effect of GONS on associative memory¹⁹. Given the demonstrated link between locus coeruleus dysfunction and decreased noradrenergic transmission in Lewy body-related cognitive impairment, this stimulation may offer greater therapeutic potential in PD-related cognitive impairment than in other neurodegenerative neurocognitive disorders such as Alzheimer’s dementia^7,20.

To ascertain the acceptability and initial feasibility of GONS for PD-MCI, we undertook a two-part study. The first part involved a wide-ranging survey to explore knowledge, attitudes, and experiences of individuals with PD in relation to non-pharmacological interventions for cognitive impairment. The second part was an open-label pilot feasibility trial involving a purposive sub-sample of five survey respondents to assess the safety, tolerability, and acceptability of GONS as a potential therapeutic intervention for individuals with PD-MCI.

Materials and methods

Results

Stage 2: Open label field study

Participant characteristics

Five participants with PD-MCI completed the open-label field study (Figure 1). The cohort included three males and two females, with a mean age of 67.2 years (range 56–82). All participants were retired or unemployed due to health-related reasons. One participant reported that their PD symptoms had a mild impact on daily functioning, three reported a moderate impact, and one reported severe functional impairment.

Figure 1. Recruitment flowchart for ONS intervention study.

Adherence to intervention protocol

All five participants adhered fully to the intervention protocol, completing three study visits and a follow-up semi-qualitative survey. There were no deviations from the protocol, and all outcome measures were completed by each participant.

Safety and tolerability

All participants completed a questionnaire assessing potential adverse effects associated with transcranial direct current stimulation (tDCS) as proposed by Brunoni. Commonly screened adverse effects included headache, neck pain, tingling, scalp pain, itching, burning, skin redness, sleepiness, trouble concentrating, and acute mood changes. No participant reported any adverse events, with all items rated as ‘absent.’

A semi-qualitative post-intervention survey using Likert scales (1–10) evaluated participants’ perceptions of the intervention. Ratings indicated a high degree of safety (mean 10/10) and acceptability (mean 9.4/10). Convenience received a slightly lower average rating (7.4/10), with travel to the study site noted as the primary inconvenience. When asked to rate the likelihood of choosing ONS as a treatment for cognitive issues in PD if found effective, participants provided a unanimous and positive response (mean 9.4/10).

Mood (BDI) and anxiety (BAI) rating scales showed no significant differences pre- and post-intervention (Table 3).

Table 3. Cognitive and neuropsychiatric outcomes measured at visit 1 (day 0) and visit 3 (day 28).

	Visit 1 (mean ± SD)	Visit 3 (mean ± SD)	Mean of differences (95% CI)	SD of differences	p value	Cohen’s d
MMSE1	27.2 ± 3.34	28.6 ± 2.6	1.4 (-0.27 to 3.07)	1.3	0.08	1.08
BDI2	15.6 ± 10.45	15.2 ± 12.52	-0.4 (-6.8 to 6.0)	5.2	0.87	0.08
BAI3	13 ± 7.75	13 ± 6.29	0.0 (-4.0 to 4.0)	3.2	>0.99	0.0
WAIS DS TRS4	26 ± 3.74	29.8 ± 4.32	3.8 (-1.42 to 9.02)	4.2	0.11	0.9
COWAT5 (letters)	49 ± 14.34	47.4 ± 16.53	-1.600 (-6.377 to 3.177)	3.847	0.41	0.42
COWAT (categories)	16.4 ± 7.23	11.4 ± 3.91	-5.000 (-14.12 to 4.124)	7.348	0.20	-0.68
CLVT-II (TL)6	31 ± 13.27	35.6 ± 12.4	4.6 (1.741 to 7.459)	2.3	0.01	2.0
Trail-making	50.35 ± 15.72	46.34 ± 23.16	-4.01 (-25 to 17)	17	0.62	-0.24

¹Mini-mental state examination; ²Beck Depression Inventory; ³Beck Anxiety Inventory; ⁴Weschler Adult Intelligence Scale, Digit Span, Total Recognition Score; ⁵Controlled Oral Word Association Test; ⁶California Verbal Learning Test II (Total Learning)

One participant mentioned that the typing component of the word association task was challenging due to Parkinsonian tremor, which impacted performance.

Preliminary evidence of efficacy

In this small feasibility trial, five participants completed the intervention, and no adverse events were reported, indicating that the intervention was well-tolerated. Descriptive analysis of cognitive outcomes suggested a potential trend toward improvement in total learning (mean difference = 4.6, 95% CI: 1.7 to 7.5, d = 2.0), though this was not statistically significant (Figure 2). Similarly, trends toward improvements in the Mini-Mental State Examination (MMSE; Figure 3) and the WAIS digit span were observed, but these did not reach statistical significance.

Figure 2. Trend to improvement in Total Learning score on California Verbal Learning Test-II.

Figure 3. Trend to improvement in Mini Mental State Examination score.

Discussion

This two-stage study provides preliminary support for the feasibility of GONS as a potential intervention for cognitive impairment in PD-MCI. Stage 1, which involved a survey examining the knowledge, attitudes, and practices of individuals with PD regarding non-pharmacological interventions, revealed that cognitive impairment is often under-recognized and undertreated in clinical practice, particularly when compared to motor symptoms. Many participants reported cognitive changes since their PD diagnosis, and following an informational video, a large majority expressed openness to exploring non-invasive neuromodulation as a treatment for cognitive symptoms. These findings align with existing literature that emphasizes the cognitive aspects of PD as a largely overlooked domain in both clinical care and research³.

The qualitative results from the survey provided deeper insights into participants' views on non-drug treatments like GONS. Several respondents emphasized that cognitive issues in PD are often treated as secondary to motor symptoms, which reflects a gap in both clinical discussions and research addressing cognitive concerns in this patient population. Participants expressed interest in the potential for non-pharmacological treatments to alleviate these symptoms. While some reported positive experiences with similar treatments, such as functional electrical stimulation for gait, others raised concerns based on prior negative encounters with electrostimulation, such as nerve conduction studies or electroconvulsive therapy. This range of feedback underscores the importance of addressing individual patient experiences and perceptions, which can vary significantly when it comes to neuromodulatory interventions.

A key concern expressed by participants was the possibility of overstimulation during GONS sessions, as well as potential withdrawal or tolerance effects. These insights suggest that future studies should incorporate comprehensive educational materials about ONS, including a Frequently Asked Questions (FAQ) section addressing common concerns about the treatment, its side effects, and its interactions with other therapies like deep brain stimulation. Clear, accessible information is essential, particularly as neuromodulation research continues to evolve. Despite concerns, participants largely recognized the potential for ONS to improve daily functioning, with many hopeful that it could help individuals with PD maintain full-time employment and engage more effectively in daily activities. These findings are critical for guiding future research and improving patient-centred care in PD.

Stage 2 of the study involved an open-label field trial to assess the feasibility and acceptability of an GONS protocol adapted from a concurrent trial in individuals with amnestic MCI. Five participants completed the trial, and all expressed strong positive impressions of the intervention. The safety and acceptability ratings were high, and all participants indicated that they would be very likely to consider GONS or a similar non-invasive neuromodulatory intervention if it was proven to be effective. One participant with more severe PD symptoms noted that their tremor interfered with fine motor tasks, such as typing, highlighting the potential impact of motor symptoms on the measurement of cognitive outcomes. This observation suggests that future protocols should account for the confounding effects of motor symptoms, such as tremor and bradykinesia, which could affect cognitive performance assessments in the PD population.

While the overall acceptability of GONS was rated at 9.4 on average by all participants, the convenience of the intervention scored lower at 7.4. The primary driver for this was felt to be the necessity of travelling to the study site to take part. Future studies utilising GONS may consider setting up participants to receive stimulation at home, either with visits from study staff or after receiving training in self-administration.

This study demonstrates that GONS is not only feasible but also well-accepted by individuals with PD-MCI. The open-label trial showed that the GONS intervention protocol is both safe and tolerable for this patient group, providing a foundation for future randomized controlled trials. Given the promising safety and acceptability data, future research could focus on refining the protocol to better mitigate the impact of motor symptoms on measurement of cognitive performance. As highlighted by a recent Cochrane review, there is a clear need for effective non-pharmacological interventions in Parkinson’s disease (PD). Pharmacological treatments offer only modest benefits and are often limited by side effects, polypharmacy, and comorbidities, which compromise their tolerability and accessibility^2,23,24.

This study is not without limitations. The self-selection bias inherent in both stages of the study may have influenced the findings, as individuals with clear cognitive symptoms or heightened interest in potential treatments may have been more likely to participate. This bias may limit the generalizability of the results. Additionally, the survey sample was relatively homogeneous and non-representative, comprising mostly individuals who are well-educated, technologically proficient, and engaged with patient organizations in Ireland and the UK. To assess potential sampling bias and the presence of predictor variable, we examined the data to assess whether there were any significant predictors of responses. No significant associations were identified, suggesting limited evidence of bias in current sample. However, considering the small sample size, the study may have been underpowered to detect subtle effects. Future studies with a larger and more diverse sample pool will be better suited to conducting stratified analyses and explore subgroup differences and potential biases.

This specific recruitment approach may not reflect the broader PD population, limiting the external validity of the results. Moreover, the survey tool itself was specifically developed for this study, and although it was modelled after existing tools in the literature, none of the instruments used were validated questionnaires. The reliance on self-reported data further introduces potential bias, particularly in the section concerning cognitive issues in PD. The subjective nature of self-reports on cognitive symptoms introduces ambiguity regarding their true presence and severity, and it is unclear whether factors other than PD may have contributed to the reported cognitive difficulties.

For Stage 2, the adaptation of the intervention protocol from a trial in individuals with amnestic MCI may have contributed to some limitations. The Mini-Mental State Examination (MMSE), used as the baseline measure of cognitive function, is primarily designed to screen for amnestic deficits and may not be the optimal tool for assessing cognitive function in PD-MCI, where impairments in executive function and visuospatial processing are more common. Future studies might consider using alternative measures, such as the Montreal Cognitive Assessment (MoCA) or the Addenbrooke’s Cognitive Examination III (ACE-III), which are more sensitive to the types of cognitive deficits seen in PD.

As stage 2 was a small, open-label study focused primarily on safety and tolerability, any evidence of efficacy must be interpreted with caution. Given the lack of control or comparator group, the possibility of a learning or repeated exposure effect exists, although there is evidence to support that the MMSE has good test-retest reliability overall²⁵. The efficacy of GONS for improving cognitive function in PD remains to be explored in larger, randomized controlled trials, stemming from this initial safety and tolerability study. Similarly, future trials might explore a longer timeframe for follow up in terms of tolerability.

Conclusion

This study underscores that cognitive impairment is an under-recognized aspect of PD, with patients expressing interest in novel neuromodulatory treatments due to limited efficacy of pharmacological options. In an open-label study, all five participants who completed GONS treatment found it safe, comfortable, and without adverse effects, with one cognitive outcome showing improvement and a non-significant increase in MMSE scores. The findings support further exploration of neuromodulatory interventions for PD-related cognitive impairment. Participant feedback highlights the need to adapt protocols for PD patients, considering motor symptoms that may hinder tasks like typing. Future randomized sham-controlled trials are necessary to assess efficacy with adequate statistical power.

Ethical approval statement

The study was conducted in accordance with the Declaration of Helsinki, and approved by the Health Sciences Ethics Committee of Trinity College Dublin (for stage 1 of the study, approval granted 1^st of November 2022, reference number 220605) and the Ethics Committee of The School of Psychology, Trinity College Dublin (for stage 2 of the study, approval granted 25^th of September 2024, approval ID SPREC102021-23).

Informed consent statement

Written Informed consent was obtained from all subjects involved in the study.

Abbreviations

The following abbreviations are used in this manuscript:

PD                 Parkinson’s disease

MCI             Mild cognitive impairment

PDD            Parkinson’s disease dementia

tDCS           Transcranial direct-current stimulation

rTMS          Repetitive transcranial magnetic stimulation

DBS             Deep brain stimulation

VNS            Vagus nerve stimulation

GONS         Greater Occipital nerve stimulation

UK               United Kingdom

LBD             Lewy body disease

MMSE        Mini mental state exam

MoCA         Montreal cognitive assessment

ACE-III      Addenbrooke’s cognitive examination III