Parkinson’s disease (PD) is a common neurodegenerative disorder caused by degeneration of midbrain dopaminergic neurons of substantia nigra (SN) producing its primary motor cardinal features (1). In addition to the motor symptoms, PD patients develop a variety of non-motor symptoms (NMSs), which significantly impair their quality of life. The NMSs consist of autonomic dysfunction, sensory symptoms, neuropsychiatric disturbances, sleep disorders, fatigue, and gastrointestinal disorders (2). Those NMSs are associated with dopaminergic and non-dopaminergic dysfunctions including serotoninergic, noradrenergic, and cholinergic systems (3, 4). Degeneration of several brainstem nuclei and their connections is responsible—at least partially—for different neurotransmitters disruption, resulting in different NMSs (3, 5).

Lewy bodies (LB) and Lewy neurites (LN) composed of alpha-synuclein are the pathological hallmarks of PD (6, 7). Alpha-synuclein pathology related to NMS was described in brain hemispheres, brainstem, spinal cord, and peripheral nervous system (5). Recent studies emphasized the importance of brainstem as the habitat of degeneration of nuclei responsible for different NMSs (8). Moreover, some NMS-related pathological changes might have specific distribution and anatomical localization in various levels of the brainstem. For example, depression and REM sleep behavior disorder are related to degeneration of pontine nuclei (5).

These reports are consistent with Braak’s proposal of pathological progression of PD, which starts caudally from the dorsal motor vagal nucleus in the medulla and then ascends in the brainstem and finally involves neocortex (9, 10). Thus, most brainstem nuclei are involved in early stages (I–III) that explain the preclinical and early emergence of NMS, while SN is involved in stage III (3, 5). Recently, Seidel et al. demonstrated the widespread of LB and LN in brainstem nuclei and fiber tracts including vestibular nuclei (4). Therefore, diagnostic tools exploring disruption of lower brainstem nuclei and related NMS are needed for early diagnosis of PD (11).

Brainstem auditory-evoked potential (BAEP) and vestibular-evoked myogenic potential (VEMP) responses represent a valuable tool for brainstem assessment as the neural pathways of both ocular VEMPs (oVEMPs) and cervical VEMPs (cVEMPs) pass through the brainstem (12). VEMPs are short latency manifestations of vestibulo-ocular reflex connecting VIII and III cranial nuclei and vestibulo-collic reflex connecting VIII and XI cranial nuclei that originate from the utricle and saccule, respectively (12). Previous studies described impaired BAEP and VEMP responses in PD patient compared to controls that were attributed to underlying brainstem dysfunction (13–16). Furthermore, impairment of these responses was related topographically to other brainstem lesions. Principally, the BAEP and oVEMP responses are affected in upper brainstem (midbrain) lesions, while cVEMP responses are involved in the lower brainstem (pontine and upper medullary) lesions (17–19). Therefore, they have a localizing value of brainstem dysfunction at different levels.

In the current study, we hypothesized that NMSs of PD related to brainstem dysfunction could be related to changes in VEMPs and BAEP. Accordingly, the aim of this study was to explore the abnormalities of BAEPs and VEMPs in patients with PD and its correlation to the motor and NMSs of PD.

Fifteen patients with idiopathic PD (12 males and 3 females) completed the clinical and audiological assessments. Their mean age was 59.20 ± 10.08 years (ranged from 35 to 70 years), and the duration of illness was 5.50 ± 2.96 years (2–10 years). The mean total UPDRS and mean disease disability (S&E scale) scores were 41.33 ± 30.20 and 68.67 ± 22.30, respectively. Patients were of variable disease stages ranged from stage 2 to 5 of H&Y scale in “Off” state with mean 2.73 ± 0.84. All patients had at least impaired domain of NMS. Sleep/fatigue (86.7%) and mood/cognition (73.3%) were the most frequent reported NMS. Detailed motor and non-motor scores of the PD patients are presented in Tables 1 and 2.

The current study investigated comprehensively the brainstem dysfunction of PD patients and its relation to motor and NMSs using BAEPs and VEMPs. It confirmed the impairment of BAEP and VEMP responses in patients with PD compared to controls that are related to clinical asymmetry of PD and its cardinal motor features. The BAEPs and cVEMPs were correlated significantly to contralateral motor UPDRS, rigidity, and bradykinesia severity, and ipsilateral dyskinesia scores. Moreover, cVEMP responses were correlated to some non-motor features including sleep/fatigue, mood/cognition, perception, and memory. BAEP and oVEMP responses showed less correlation to the motor and non-motor features. Consequently, it reflects brainstem pathology among PD patients at different levels and highlights the asymmetry of these changes.

In this study, we separated the vestibular responses to ipsilateral and contralateral responses according to CMA side. This approach could reflect the expected asymmetrical brainstem pathology of PD (25). Previous studies confirmed the asymmetry of cardinal motor symptoms that often persists through the course of the disease. This clinical asymmetry is related to asymmetrical degeneration of dopaminergic neurons of SN, striatal dopaminergic receptors, and their cortical connections (8, 25). Furthermore, the clinical asymmetry could be related to NMSs including pain, fatigue, depression, and cognitive dysfunction. This could be explained by asymmetrical underlying pathological changes and the asymmetrical dopaminergic deficiency that contributes to some non-motor features along with other neurotransmitters (8, 25).

The high-frequency hearing impairment associated with PD was detected in the current work similar to prior studies (26–28). Impaired hearing in PD was attributed to peripheral auditory dysfunction (27, 28) and abnormal central auditory processing (29, 30). Likewise, abnormal BAEPs among PD patients were demonstrated inconsistent to previous studies, denoting brainstem auditory conduction delay (26, 31, 32).

Additionally, both the oVEMPs and cVEMPs were affected in the studied PD subjects. The findings of the present study agreed with de Natale et al. who reported that the frequency of alteration of VEMPs in PD patients was 83.3% when considering the combined set of cVEMP, oVEMP, and masseter VEMP responses, with absence being the prevalent alteration in PD (14). Similarly, Pollak et al. showed unilaterally absent VEMP responses in 20 (37%) of PD patients and bilaterally absent responses in 4 patients (7.4%) (33). Lower amplitudes of cVEMPs among PD patients were reported by another study (34). However, a recent study by Pötter-Nerger et al. reported abnormal oVEMPs in mild to moderate PD patients with preserved cVEMPs that were attributed to methodological differences (16). These findings emphasize the extensive brainstem dysfunctions at different anatomical levels (35).

Auditory and vestibular dysfunctions in PD could be explained by different mechanisms. Modulatory effect of dopamine on the excitability of vestibular nuclei is one of these mechanisms (34). Disrupted interconnections of vestibular nuclei with degenerated other brainstem nuclei by PD pathology especially dorsal raphe nuclei is another mechanism (36, 37). Furthermore, direct disruption of vestibular nuclei by PD pathological changes produces these vestibular abnormalities as recently reported (4).

Remarkably, the current study defined correlation of the main motor features of PD with BAEPs and cVEMPs rather than oVEMPs that might be explained by the midbrain and pontine pathological changes and non-involvement of vestibulo-ocular pathways in the pathophysiology of these features. Rigidity and bradykinesia were related to BAEP and cVEMP responses, while tremor was not. The correlations were mainly to the responses contralateral to CMA side, which is consistent with asymmetric nature of PD pathology in SN and their connections (25).

Tremor has different pathophysiology compared to rigidity and bradykinesia and characterized by involvement of the cerebellum–thalamocortical circuit in its pathogenesis (37). This explains the lack of correlation with auditory and vestibular responses. Moreover, recent animal studies demonstrated that brainstem structures such as pontine nuclei and locus coeruleus are involved in the pathophysiology of levodopa-induced dyskinesia (LID) (38). This could explain the correlation seen in the present study between LID and cVEMP wave latencies.

Although the correlation of BAEPs, oVEMPs, and cVEMPs were mainly to one side, yet no differences existed in all tests between the two sides. The medication state of the subjects could be the explanation. All PD subjects underwent the tests during “medication on state” that masked the abnormalities between the two sides as recently reported by Pötter-Nerger et al. (34).

In contrast, previous studies reported a lack of correlation with clinical motor scores (14, 33, 34). Nonetheless, de Natal et al. defined the progression of VEMP abnormalities with increased stage of the disease (14). They used the mean values of VEMP responses on both sides, not about the CMA side, thus underestimating the potential asymmetry that could ameliorate abnormalities. Moreover, differences in experimental conditions during testing, age differences between cases and controls, and different clinical characteristics of recruited patients could explain the inconsistency of results of different studies that addressed the vestibular functions in PD.

Few prior studies investigated the correlation of auditory and vestibular responses with individual NMSs of PD (14, 33). In the current study, sleep and mood domains’ severity demonstrated correlation with cVEMP responses that could be related to associated dysfunction of different pontine nuclei such as locus coeruleus, raphe nucleus, and pedunculopontine nucleus (3). Previous studies reported localizing function of cVEMP responses of pontine lesions associated with different other diseases (17–19). Recently, de Natale et al. found a direct correlation between VEMP changes and REM sleep behavior disorder and postural instability (14). The same study reported a lack of correlation with depression (14), while a study by Pollack et al. reported a correlation of cVEMP with depression and antidepressants use (33).

Unsurprisingly, GIT domain that includes hypersalivation, dysphagia, and constipation had minor correlations. This could be attributed to the high contribution of peripheral pathological changes in GIT and related nerve supply (alpha-synucleinopathy) along with brainstem changes (39–41). Furthermore, GIT symptoms are attributed to dysfunction of the dorsal motor nucleus of the vagus, nucleus ambiguous, and nucleus of a solitary tract located in the medulla (3) that is poorly localized by VEMP responses (13). Correlation of GIT symptoms’ severity to contralateral BAEP latency could be explained by associated advanced disease (42). Similarly, the contribution of sacral spinal cord alpha-synuclein deposition to urinary symptoms explains limited correlation to vestibular responses (43). Likewise, correlation of other non-motor features poorly linked to brainstem dysfunction such as cardiovascular symptoms, attention/memory, and perception/hallucination could be explained by associated disease severity (42). Cognitive deficits of PD are correlated to higher order auditory processing in subcortical–cortical pathways evaluated by event-related potentials (P300) (29, 44).

The current study has different limitations. These limitations include a small number of recruited sample, including patients with different disease severity and stages, and the use of NMSS subscores that describe different NMSs for correlations with vestibular responses. Thus, further, more specific tools for each NMS are required for correlation with the electrophysiological assessment of the larger number of PD patients.

In conclusion, the current study confirms the auditory and vestibular abnormalities among PD patients that reflect brainstem pathological changes. It also correlates these abnormalities to some motor and non-motor features of the disease, providing a localizing tool for associated brainstem dysfunction. Furthermore, abnormal vestibular potentials are related to disease severity and stage and might respect clinical and pathological asymmetry. However, further studies are warranted to reproduce the correlation of VEMP responses to the individual motor and NMSs of PD and their value as potential biomarkers at different stages of the disease, especially the early PD.

The ethical committee of the Faculty of Medicine, Ain Shams University has allowed doing this study. The authors assert that all procedures contributing to this work comply with the ethical standards of the relevant national and institutional committees on human experimentation and with the Helsinki Declaration of 1975, as revised in 2008.

AS and DH: research idea and conception, data acquisition, data analysis and interpretation, and manuscript writing and reviewing. HE, EM-H, and JS-P: reviewing the manuscript. MS: data analysis and manuscript reviewing. OA-C: data output analysis and manuscript writing and reviewing.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.