Oral health and oral microbiota composition in patients with Parkinson’s disease: a case‒control study

ABSTRACT

Introduction

Parkinson’s disease (PD) is associated with motor and non-motor dysfunctions that may influence oral homeostasis. An altered microbiota has been linked to PD, yet oral microbial changes remain poorly characterized. This case‒control study aimed to determine if the oral microbiota, the oral health status, and salivary secretion differ between patients with PD and matched controls.

Materials and methods

Forty participants (20 with PD and 20 controls) underwent standardized oral examinations, neurological scoring (Hoehn & Yahr; UPDRS II–III), and measurements of unstimulated (UWS) and stimulated whole saliva (SWS) flow rates. Microbial profiling of SWS samples and dorsal tongue scrapes was performed using long-read 16S rRNA gene sequencing (V1–V8). Alpha/beta diversity and differential abundance analyses were performed.

Results

Patients with PD exhibited significantly lower UWS flow rates, higher plaque scores, and higher prevalence of xerostomia and drooling. Dental status and gingival inflammation did not differ between the groups. Microbial richness and Shannon diversity were lower in tongue scrapes from the PD group, but not in saliva samples. Veillonella, Prevotella and Streptococcus dominated the microbiota in both groups. Fusobacterium periodonticum and pseudoperiodonticum were consistently more enriched in both saliva and tongue microbiota of patients with PD.

Conclusion

Patients with PD have lower UWS flow rates and a shift toward a more anaerobic, biofilm-adapted oral ecosystem, particularly on the tongue. The tongue microbiome may potentially serve as a sensitive biomarker of PD-related dysbiosis.

Key messages

• Patients with Parkinson’s disease exhibit reduced unstimulated salivary flow, increased plaque accumulation, and more frequent symptoms such as xerostomia and drooling compared with controls.

• The oral microbiota of patients with Parkinson’s disease, particularly on the tongue, shows reduced diversity and a shift toward an anaerobic, biofilm-adapted ecosystem with enrichment of specific taxa such as Fusobacterium species.

Introduction

Parkinson’s disease (PD) is the second most common neurodegenerative disorder, primarily characterized by motor symptoms such as tremors, rigidity, and bradykinesia, along with a wide range of debilitating non-motor manifestations including cognitive impairment, depression, gastrointestinal dysfunction, and oral health problems [1–3].

Beyond its core neurological features, PD is accompanied by a range of oral health challenges, which arise from the combined influence of motor symptoms, non-motor symptoms, and adverse effects of pharmacological treatment. Motor impairments such as tremors, rigidity, and bradykinesia limit patients’ ability to perform effective oral hygiene, increasing their susceptibility to plaque accumulation, dental caries, and periodontal disease [3–7]. Dysfunction of orofacial and oropharyngeal structures, marked by impaired muscular coordination, reduced tongue mobility, and dysphagia, further impairs chewing efficiency, swallowing function, and salivary clearance [8–12]. In addition, cognitive decline, depression, and apathy, all common non-motor features of PD, may reduce patients’ motivation and ability to maintain daily oral hygiene and to seek timely dental care [13]. Prolonged dopaminergic therapy has also been linked to worsened xerostomia and hyposalivation, further heightening the risk of dental disease [11,14,15]. Persons with PD also appear to have a greater burden of periodontitis, but evidence does not indicate a reciprocal relationship between PD and periodontitis risk [5,6,16].

Emerging evidence suggests that alterations in the microbiota across mucosal surfaces, including the gut, oral cavity, and nasal passages, may play a significant role in the pathogenesis of PD [17–22]. Particularly the gut microbiota can influence the central nervous system through the neural, endocrine, and immune pathways of the microbiota–gut–brain axis [23]. Strong associations between gut microbiota dysbiosis and PD have been consistently documented [23,24].

Dysbiosis, defined as an imbalance or disruption in the microbial community, is increasingly associated with systemic diseases such as obesity, diabetes, cardiovascular disorders, and neurodegenerative conditions. This underscores the critical role of microbial ecosystems in maintaining systemic health and modulating disease development [25,26]. Despite advances in microbiome research, a substantial gap remains in understanding the nature and consequences of disruptions of the oral microbiota in PD. Most studies have focused on gut microbiota changes, consistently identifying differences in microbial diversity and specific bacterial abundances between PD and control groups [20,22]. By contrast, the oral microbiome has received comparatively little attention, although the oral cavity is a gateway to the gastrointestinal tract and is strongly linked to systemic inflammation and disease.

Preliminary investigations have reported oral dysbiosis in PD in terms of a higher prevalence of pathogenic bacteria and altered inflammatory profiles in oral tissues [27,28]. Streptococcus species appear to be enriched in PD, particularly S. mutans and S. anginosus [17,28–31]. The enrichment of other species, such as S. sobrinus, has also been observed but is considered less relevant [29]. Families including Lactobacillaceae and Prevotellaceae are also commonly reported to be enriched [17,20,25,27,29,31] as well as Veillonellaceae [17,28]. Notably, only two studies have identified Porphyromonas gingivalis as a primary pathogen [19,21].

These previous findings are limited by methodological variability and a lack of systematic exploration of the relationship between oral health and PD pathology. To address these gaps, this case‒control study comprehensively evaluated the oral health status and functions and the microbiota composition in whole saliva and dorsal tongue scrape samples from participants with PD and without PD, including identification of potential differences in bacterial abundances and alpha and beta microbial diversity. We hypothesized that patients with PD would have poorer oral health and present an oral microbial profile that differs from that of matched controls across two distinct oral sites.

Materials and methods

Study design, setting and participants

This clinical case‒control study was conducted from May 2021 to October 2022 at the Department of Odontology, University of Copenhagen, Denmark, in collaboration with the Department of Neurology, Bispebjerg Hospital, Copenhagen, Denmark.

A total of 40 participants were enrolled: 20 patients diagnosed with PD and 20 age- and sex-matched controls without PD. Participants were recruited as outpatients from the Department of Neurology, Bispebjerg Hospital, where diagnoses were established by experienced neurologists. All patients with PD underwent detailed neurological evaluations to determine disease severity and functional status. Disease progression was assessed using the modified Hoehn and Yahr (H&Y) scale [32], which ranks PD severity from stage 1 (mild symptoms) to stage 5 (severe disability). Disease severity and motor performance were further assessed using the Unified Parkinson’s Disease Rating Scale (UPDRS) [33]. Specifically, Part II (Motor Experiences of Daily Living) is a patient-reported measure evaluating speech, swallowing, hygiene, salivary secretion, and the ability to cut food, among other daily motor functions. Part III (Motor Examination) is a clinician-scored assessment focusing on speech volume, facial expression, rigidity, bradykinesia, and other motor features of PD. Together, these scales provide complementary quantitative and qualitative measures of PD severity and its impact on motor function and daily living.

Participants in the control group were initially recruited among the spouses of patients with PD to ensure similarity in diet, lifestyle habits, oral hygiene habits, and socioeconomic background. To broaden recruitment, additional participants without PD were enrolled via flyers distributed in the hospital, university, and public libraries.

Eligibility criteria for both groups included the ability to comprehend and comply with study procedures and the ability to attend the study site independently. The exclusion criteria were cognitive impairment, a diagnosis of Sjögren’s disease, the presence of implanted electronic devices such as pacemakers or deep brain stimulators, and prior head or neck radiotherapy.

The study was conducted in accordance with the Declaration of Helsinki. Ethical approval was granted by the Regional Committee on Health Research Ethics of the Capital Region (H-20047464) and the Danish Data Protection Agency (514-0510/20-3000). The study was prospectively registered at ClinicalTrials.gov (NCT05356845). Written informed consent was obtained from all participants after they received both oral and written information about the study.

Clinical examination and interview

All participants were examined by a single investigator (co-author SB) at the Department of Odontology, Faculty of Health and Medical Sciences, University of Copenhagen.

A standardized interview and questionnaire were administered to collect information on general health, oral health and dental habits, daily medication use, and use of tobacco. Xerostomia and drooling were further evaluated with validated questionnaires to assess the severity of symptoms. Xerostomia was scored on a four-point scale: (0) no sensation of dry mouth, (1) mild sensation of dry mouth, (2) severe sensation of dry mouth, and (3) annoying sensation of dry mouth that significantly impairs speech [34]. Drooling was assessed by a questionnaire rating both frequency and quantity, yielding a total score from 2 (no drooling) to 9 (severe drooling). Scores above 3 were classified as indicative of clinically relevant drooling [35].

Saliva samples were collected between 08:00 and 12:00 to minimize circadian variation in secretion and composition. The collection was standardized to occur one hour after Levodopa administration. The participants were instructed to refrain from eating, drinking, smoking, chewing gum, or toothbrushing for at least one hour before sampling. Unstimulated whole saliva (UWS) and paraffin-chewing stimulated whole saliva (SWS) were collected for 5 min using the draining method [36,37]. Hyposalivation was defined as UWS flow rates ≤ 0.1 ml/min and SWS flow rates ≤ 0.7 ml/min. SWS samples were used for microbiological analyses. UWS flow rates of 0.3–0.4 ml/min and SWS flow rates of 1.5–2.0 ml/min were considered within the normal range. All saliva samples were kept on ice immediately after collection, aliquoted, and stored at –80 °C until analysis.

For tongue scrape sampling, a sterile single-use brush was drawn over the entire dorsum of the tongue in vertical strokes from posterior to anterior for approximately 15 s. The collected material was transferred into a sterile tube containing Tris-EDTA (TE) buffer, and the brush was thoroughly rubbed against the tube wall to release the remaining material. The samples were kept on ice and stored at –20 °C until analysis.

SWS was selected because it provides a representative overview of the entire oral cavity, yields a sufficient sample volume, and is compatible with the sampling approaches used in many other oral microbiome studies [38–40]; SWS also captures greater microbial shedding than UWS. Tongue scraping was included in addition to whole saliva, as it represents a niche with distinct microbial specificity [41].

The clinical oral examination included assessment of periodontal pocket depth, gingival index, and plaque index, performed on six index teeth (16, 12, 24, 44, 32 and 36). The plaque index and gingival index were recorded at four sites per tooth (three facial, one lingual) using a manual periodontal probe [37,42]. Periodontal pocket depth was defined as the distance from the free gingival margin to the base of the pocket, measured in millimeters; sites with a pocket depth ≥ 5 mm were additionally recorded at six sites per tooth (three facial, three lingual) [37,42,43]. Decayed, missing, and filled teeth (DMF-T) and surfaces (DMF-S) were registered according to WHO criteria [43].

Masticatory efficiency was assessed by recording the time required to chew and swallow a standardized 10 g piece of apple [44].

Swallowing capacity was determined using the ‘Timed Water Swallow Test’, expressed as milliliters swallowed per second [45].

Olfactory function was assessed with the Brief Smell Identification Test (BSIT), a 12-item version of the University of Pennsylvania Smell Identification Test [46]. Scores ranged from 0 to 12, with higher scores indicating better olfactory recognition. For each item, a microcapsule containing a common odor was scratched and presented under the participant’s nose, and the participant selected the correct odor from four options (e.g. motor oil, garlic, rose and lemon).

Microbial marker gene analysis

DNA extraction, library preparation, and sequencing of the 16S rRNA gene (variable regions V1–V8, bV18-A) were performed by DNASense Aps [47].

DNA extraction and library preparation

DNA was extracted from chewing-stimulated saliva and tongue scrape samples obtained from all 40 participants using the DNeasy Blood and Tissue Kit (Qiagen, Germany) with minor protocol modifications. The DNA concentration was measured using a Nanodrop spectrophotometer and a Qubit dsDNA BR Assay Kit (Thermo Fisher Scientific, USA), and the integrity was verified with the Tapestation 2200 (Agilent, USA).

Amplicon libraries targeting the 16S rRNA V1–V8 regions (bV18-A) were generated from up to 25 ng of DNA by PCR (50 µL reactions containing 0.2 mM dNTPs, 0.01 U Platinum SuperFi DNA Polymerase, and 500 nM primers). The cycling conditions were: 98 °C for 3 min; 25 cycles of 98 °C for 30 s, 62 °C for 20 s, and 72 °C for 2 min; followed by a final extension at 72 °C for 5 min. The primers used contained 24-nucleotide barcodes and included (8F) AGRGTTYGATYMTGGCTCAG and [1391 R] GACGGGCGGTGWGTRCA [48–51]. The PCR products were purified using CleanNGS beads (3:5) and eluted in 25 µL of water. Libraries were subsequently prepared with the SQK-LSK114 kit (Oxford Nanopore, UK) using 500 ng of input DNA, bead-based cleanups, quantification with the Qubit HS assay, and quality assessment with the Tapestation.

Sequencing and bioinformatics

Sequencing was performed on a PromethION R10.4.1 flow cell with MinKNOW v24.06.15. The reads were basecalled and demultiplexed using Dorado v7.4.14 (super-accurate model, custom barcodes). Quality filtering with Filtlong v0.2.1 retained reads 320–2000 bp in length with Phred scores >17. Reads were mapped to GTDB r220 SSU representatives (bac120, ar53) using minimap2 v2.24 [52] and processed with samtools v1.14 [53]. The exclusion criteria included alignments with >5% length deviation, bacterial SSU <650 bp, archaeal SSU <450 bp, operational taxonomic units (OTUs) <0.1% relative abundance, or <10 reads per sample.

Bioinformatic analyses were carried out in R v4.4.1 (RStudio IDE 2023.12.1.402) using ampvis2 (2.8.7), tidyverse (2.0.0), seqinr (4.2.36), ShortRead (1.60.0), and iNEXT (3.0.1). Samples with fewer than 10,000 mapped reads were excluded. Successful sequencing was achieved for 39 of 40 saliva samples and all 40 tongue scrape samples [54–56].

Statistics

Sample size calculation: A previous study [44] reported that individuals with PD had a 50% risk of experiencing orofacial symptoms, compared with 10% among controls. Based on a power of 80% and a significance level of 0.05, a sample size of 19 participants per group was required. Accordingly, we included 20 participants with PD and 20 control participants [57].

Statistical analysis: Statistical analysis was conducted using IBM SPSS Statistics 29. Initially, descriptive statistics were employed to analyze the characteristics of the study population, including age, sex and number of teeth. For the patients with PD, the analysis also encompassed disease status, including disease duration, H&Y staging and UPDRS II and III assessments. As the data were not normally distributed, non-parametric analyses were conducted, including the Mann‒Whitney U test for continuous variables and the χ2-test for categorical variables. The small number of participants made it unfeasible to adjust for multiple explanatory variables in regression analyses.

The adonis2 function in the R vegan package [58] was used for PERMANOVA analysis. P-values were FDR-adjusted for multiple hypothesis testing. ANCOM from the package ANCOMBC was used (v1.6.4) to analyze the composition of the microbiome to obtain significantly different taxa between the two groups, while controlling for age as a confounder [59]. For each amplicon sequence variant (ASV), the fraction of samples showing significant differential abundance between groups was calculated. ASVs were reported if significant enrichment was observed in more than 60% of the samples (fraction > 0.6), with additional annotations for higher fractions (0.7, 0.8, and 0.9). This statistical framework accounts for inter-sample variability while controlling for age.

Statistical significance was set at p < 0.05.

Results

Demographic and clinical characteristics of the participants

The demographic characteristics of the study participants are summarized in Table 1. The two groups were comparable in terms of age, sex distribution, and number of teeth, including both natural teeth and fixed replacements.

Table 1.

Demographic and clinical characteristics of the PD and control groups. Values are given in median (range) and in numbers of patients (%).

	PD (n = 20)	Controls (n = 20)	P-value
Age in years	68.5 (35–80)	67 (35–83)	N.S.
Male	6 (30%)	7 (35%)	N.S.
Female	14 (70%)	13 (65%)
Duration of PD	10.5 (3–23)	–
Disease score H&Y	3 (1–4)	–
Disease score UPDRS II	12 (5–23)	–
Disease score UPDRS III	19 (11–30)	–
Number of natural teeth	27 (3–28)	26 (22–28)	N.S.
Number of teeth, including fixed replacement, e.g. dental implants and bridges	27 (18–31)	27 (22–31)	N.S.
Plaque index score	0.25 (0–1.33)	0 (0–0.55)	0.004
Gingival index score	0.5 (0–1.71)	0.32 (0–1.42)	N.S.
DMF-T total score	14 (2–28)	15 (3–24)	N.S.
Decayed teeth	0 (0–4)	0 (0–1)	N.S.
Missing teeth	1 (0–25)	2 (0–6)	N.S.
Filled teeth	12 (1–18)	13.5 (3–23)	N.S.
DMF-S total score	43 (1–128)	52 (5–89)	N.S.
Decayed tooth surfaces	0 (0–10)	0 (0–2)	N.S.
Missing tooth surfaces	5 (0–116)	10 (0–30)	N.S.
Filled tooth surfaces	36 (2–69)	36.5 (5–89)	N.S.
UWS (ml/min)	0.24 (0.00–0.8)	0.42 (0.04–1.3)	0.009
SWS (ml/min)	1.52 (0.10–3.30)	1.80 (0.66–3.94)	N.S.
Hyposalivation, UWS ≤ 0.1 ml/min	6 (30 %)	1 (5 %)	0.037
Hyposalivation, SWS ≤ 0.7 ml/min	5 (25%)	1 (5%)	N.S.
Assessment of xerostomia severity:
No feeling of dry mouth	8 (40%)	16 (80%)	0.024
Slight feeling of dry mouth	7 (35%)	4 (20%)
Severe feeling of dry mouth	2 (10%)	0
An annoying feeling of dry mouth that makes	2 (10%)	0
speech difficult
-Missing	1 (5%)
Drooling score, subjective	4 (2–7)	2 (2–4)	<0.001
Masticatory efficiency of 10 g apple chewing, seconds	25 (14–80)	20 (10–31)	0.023
Swallowing capacity, mL water pr. second	10.7 (1.8–25)	21.4 (9.4–50)	<0.001
Olfactory Testing, B-Sit score	5.5 (0–10)	10 (8–12)	<0.001
Regular dentist/dental hygienist visits:
At least once a year for a dental check-up	20 (100%)	18 (90%)
2–4 times in the past 5 years	0	2 (10%)	N.S.
1–2 times in the past 5 years	0	0
No visits in the past 5 years	0	0

PD: Parkinson’s disease.

H&Y: Modified Hoehn and Yahr scale.

UPDRS: Unified Parkinson’s Disease Rating Scale. II: Motor experience of daily living; III: Motor function.

DMF-T: Decayed, missing, filled teeth.

DMF-S: Decayed, missing, filled surfaces.

UWS: unstimulated whole saliva flow rate.

SWS: chewing-stimulated whole saliva flow rate.

B-Sit: Brief Smell Identification Test.

In the PD group, the median disease duration was 10.5 years (range 3–23), the median H&Y stage was 3 (1–4), and the median UPDRS II and III scores were 12 (5–23) and 19 (11–30), respectively.

The medication profile clearly differed between the two groups. All participants in the PD group were on disease-specific medications, including levodopa, dopamine agonists, anticholinergics, monoamine oxidase-B inhibitors, catechol-O-methyltransferase inhibitors, and monoamine oxidase inhibitor (amantadine). Intake of medications targeting non-motor disease manifestations, including analgesics, laxatives, proton pump inhibitors, hypnotics, sedatives and antidepressants, was also common. In the non-PD group, the most common medication intake included antihypertensives, antithrombotics, lipid-lowering agents, vitamin D supplements and hormone replacement therapy.

There was no statistically significant difference in tobacco use between the groups, with 2 participants in the PD group and 4 participants in the control group reporting current smoking. All participants, except for one in each group, reported using fluoride toothpaste daily, and none reported using an antiseptic mouthwash. No participants used posterior removable dentures. However, one participant in the control group had a partial removable denture limited to the anterior teeth. The dental attendance patterns were similar between the two groups. All patients with PD and 90% of the participants in the control group reported visiting a dentist at least once annually over the past five years.

The two groups differed significantly with respect to the oral and salivary findings. Hyposalivation (UWS ≤ 0.1 mL/min) was more prevalent in the PD group than in the control group (30% vs. 5%, p = 0.037). The median UWS flow rates were significantly lower in the PD group (0.24 mL/min, range 0.00–0.80) than in the control group (0.42 mL/min, range 0.04–1.30; p = 0.009). No significant differences were found for SWS flow rates.

Xerostomia was significantly more prevalent in the PD group than in the control group (60% vs. 20%, p = 0.024). In the PD group, 20% reported either severe xerostomia or annoying xerostomia affecting speech.

In terms of oral hygiene, the median plaque index score was higher in the PD group (0.25, range 0–1.33) than in the control group (0, range 0–0.55; p = 0.004), while the gingival index score did not differ significantly. No significant difference was observed between the groups in periodontal pocket depth measurements. In the PD group, 12 participants had no periodontal pockets ≥5 mm, 5 participants had 1–5 sites with pockets ≥5 mm, and 1 participant had more than 5 sites with pockets ≥5 mm. In the control group, 10 participants had no periodontal pockets ≥5 mm, 7 participants had 1–5 sites with pockets ≥5 mm, and 2 participants had more than 5 sites with pockets ≥5 mm. Similarly, measures of dental caries experience (DMF-T and DMF-S scores and their subcomponents) were comparable between groups (all p > 0.05).

The median drooling scores were significantly higher in the PD group than in the control group (4 (range 2–7) vs. 2 (range 2–4); p < 0.001). Orofacial function was poorer in patients with PD, as they required significantly more time for mastication and swallowed fewer milliliters of water per second. Also, the olfactory function, assessed with the B-SIT test, was significantly poorer in the PD group than in the control group (median 5.5 (range 0–10) vs. 10 (range 8–12); p < 0.001). Analysis of the relationship between UWS and SWS flow rates and olfactory performance revealed that B-SIT scores were negatively associated with UWS and SWS flow rates in the PD group (R² = 0.059 and R² = 0.083), but the opposite was true in the control group (R² = 0.070 and R² = 0.11) (Figure 1). However, none of the associations reached statistical significance, indicating that UWS and SWS flow rates were not strong predictors of olfactory performance in any of the groups.

Figure 1.

Association between unstimulated whole saliva (UWS) flow rate and olfactory function (B-SIT score) in participants with Parkinson’s disease (PD, blue) and controls (red). Each dot represents an individual participant. Regression lines with 95% confidence intervals are shown for the PD (blue) and control (red) groups. Linear regression equations and R² values are displayed. Neither association reached statistical significance.

Measures of microbial diversity within and across samples

In total, 587 unique taxa were identified at the species level, with individual samples yielding between 22,463 and 119,774 high-quality reads, and a cumulative total of 7,419,634 reads.

Table 2 presents the total number of unique sequences (operational taxonomic units, OTUs) and the Shannon diversity index for the PD and control groups, based on pooled saliva and tongue scrape samples. The three predominant bacterial genera identified in both saliva and tongue scrape samples were Veillonella, Prevotella and Streptococcus, accounting for 45% of total reads in the saliva and 59% in the tongue scrape. Figure 2 depicts the relative abundance of the top 20 bacterial genera in the saliva and tongue scrape samples based on the OTUs.

Table 2.

Values are presented as median (range). Richness/OTU refers to the total number of unique sequences (operational taxonomic units, OTUs) per sample, calculated using iNEXT. Shannon refers to the sample Shannon diversity index, also calculated using iNEXT.

	All	PD	Controls
Richness/OUT	79 (27–153)	107 (57–153)	62.5 (27–99)
Shannon	3.45 (2–4)	3.78 (3–4)	3.16 (2–4)

Figure 2.

Relative abundance of the top 20 bacterial genera detected in tongue scrape and saliva samples. Abundances are expressed as percentages of total sequences per sample type. Prevotella, Veillonella, and Streptococcus were among the most dominant genera across both sample types.

When pooling across sample types, the richness/total number of OTUs did not differ significantly between the PD and control groups. The higher median richness observed in the PD group in the pooled analysis (Table 2) appears to be driven by a higher total number of OTUs in the saliva samples specifically.

Assessment of the alpha diversity (Figure 3) displayed a significant group difference in terms of lower richness in tongue scrapes from the PD group than from the control group (p = 0.008), whereas no difference was observed in saliva samples (p = 0.75). Similarly, Shannon diversity was significantly lower in tongue scrapes from the PD group than in those from the control group (p = 0.0012), with no significant difference in saliva samples (p = 0.3).

Figure 3.

Alpha diversity measures for each sample site for participants with PD and without PD (non-PD).

Beta diversity was assessed using Bray–Curtis dissimilarity. For both SWS and tongue scrape samples, PD and control samples showed visual separation when plotting the first two principal coordinates (Figure 4).

Figure 4.

Beta diversity of samples across the different sampling sites for participants with PD and without PD (non-PD).

To further examine the impact of clinical and demographic factors on beta diversity, a PERMANOVA was performed, assessing the effects of sampling site, disease status, age, SWS, and sex. All the factors significantly influenced beta diversity. While the sampling site explained 7.9% of the variance in the microbiome composition, the PD status explained 5.5% of the variance (Table 3), supporting the visual separation in principal coordinates. Age, SWS flow rate, and sex explained approximately 2% of the variance.

Table 3.

Proportion of variance in beta diversity explained by different factors in variation in microbial community composition. R² = proportion of variance explained by each factor; P-adj = permutation-based p-value indicating the significance of the factor’s effect.

Factor	R2	P-adj
Site	0.079117	0.001
Disease	0.055033	0.001
Age	0.020206	0.001
SWS	0.019769	0.001
Sex	0.019595	0.001
Patient ID	0.594776	0.001

Analysis of the bacterial community composition revealed distinct differences in the relative abundance and taxonomic distribution between the PD and control groups in both saliva and tongue-scrape samples (Figure 5).

Figure 5.

Group-wise mean genus relative abundance across the two sites, with samples grouped by PD and non-PD control groups. Only the top 20 genera are shown.

The results of the differential abundance analysis performed using ANCOM are summarized in the volcano plots (Figure 6). In both saliva and tongue scrape samples, two taxa belonging to the genus Fusobacterium – Fusobacterium periodonticum and Fusobacterium pseudoperiodonticum – were identified as significantly more enriched in the PD group than in the control group.

Figure 6.

Differential abundance analysis of saliva and tongue scrape microbiota in PD versus non-PD controls. Volcano plots display the log fold change (lfc, x-axis) against the significance level (log-transformed p-values, top panels) and test statistics (W, bottom panels) for each taxon. Each point represents an operational taxonomic unit (OTU). Taxa highlighted in green were identified as significantly different between groups, with a larger point size reflecting a higher prevalence (fraction).

Discussion

The aim of this study was to evaluate and compare the oral health and functions and oral microbiota composition of patients with PD and age- and sex-matched controls without PD. By integrating neurological assessments, clinical oral examinations, salivary flow measurements, and microbial community profiling of both stimulated saliva and tongue-scrape samples, this study aimed to determine whether patients with PD exhibit characteristic changes in oral health status and the composition of the oral microbiota.

The main findings revealed that patients with PD had significantly lower UWS flow rates and higher plaque index scores than participants in the control group, whereas they did not differ with respect to SWS flow rates, dental health status and gingival inflammation. A persisting decrease in unstimulated whole saliva secretion promotes bacterial adhesion and retention of dental plaque, as oral clearance is impaired and saliva may become more viscous. This can result in increased caries activity and gingivitis. In this study, however, patients with PD were able to maintain a dental health condition compatible with that of the participants in the control group, most likely due to the maintenance of a normal stimulated salivary gland secretion and regular dental follow-up visits.

The higher plaque index score in the PD group can also be explained by motor impairment, rigidity, and tremors that reduce toothbrushing efficiency [9,11,57]. The overall difference likely reflects functional limitations rather than poor adherence, as PD participants demonstrated commitment to professional dental care. Maintaining frequent dental visits appears to mitigate risk [13,60], emphasizing the importance of continued preventive care for patients with PD.

The two groups did not differ in demographic factors or in major dental variables, including dental habits. This strengthens the validity of the observed differences, indicating that the microbial and physiological variations likely reflect disease-related mechanisms rather than disparities in lifestyle habits, oral hygiene procedures or socioeconomic background. The inclusion of spouses as controls represents a major methodological strength, as a shared diet, environment, and lifestyle minimize external confounding factors and highlight PD-specific physiological alterations. However, the PD group exhibited a higher overall medication burden, both in terms of the total number of medications used and the diversity of drug categories represented.

The findings in the PD group of lower unstimulated flow rates, a higher prevalence of UWS hyposalivation, a higher prevalence of xerostomia, and more severe xerostomia indicate that medication intake is the cause. The effect of medications on salivary secretion is complex, as a drug may simultaneously interact with the salivary secretory reflex at the level of the central nervous system and at the level of the neuroglandular junction [61,62]. Anticholinergic agents act on the central nervous system and at the peripheral level (neuroglandular junction) by reducing the release of the neurotransmitter acetylcholine and therefore particularly cause xerostomia and salivary gland hypofunction [63,64]. Muscarinic receptors (primarily M1 and M3) are the main targets in salivary glands, and their stimulation by acetylcholine is crucial for salivary secretion. Dopaminergic agents, sedatives, hypnotics, and antidepressants may also inhibit salivary secretion, mainly by acting centrally. Medication usually affects the unstimulated salivary gland secretion, whereas the salivary glands retain functional capacity upon stimulation, e.g. taste and chewing. However, both unstimulated and stimulated salivary gland secretion may be reduced, e.g. in the case of competitive inhibition of receptors involved in salivary secretion, dosage, and polypharmacy [63–66]. Cognitive decline, bradykinesia, and reduced spontaneous swallowing frequency may also contribute to a lower UWS flow rate by reducing reflexive activation of salivary nuclei [67,68]. In this study, the findings indicate that the secretory salivary gland tissue remains intact, while neural control of baseline secretion is impaired.

From a methodological perspective, the use of SWS samples for microbiota analysis is advantageous because it provides sufficient volume and comparability with previous studies. Furthermore, since some participants had low UWS secretion or hyposalivation, focusing on SWS ensured consistency and data reliability across groups. The correlation analysis between B-SIT olfactory performance and salivary flow revealed a weak but opposite trend in PD and controls: poorer olfactory function tended to coincide with lower UWS and SWS flow rates in PD, though not statistically significant. This pattern may suggest that hyposalivation and impaired olfaction share overlapping neurodegenerative or autonomic pathways. Olfactory decline is among the earliest non-motor symptoms of PD [69] and may reflect broader central nervous system involvement that also affects salivary nuclei. Together, these observations highlight the clinical interconnection between sensory and autonomic dysfunction in PD.

One of the most striking findings was the significantly lower microbial richness and Shannon diversity in tongue scrape samples from patients with PD. Reduced microbial diversity on the tongue surface may result from ecological shifts linked to impaired oral clearance and less frequent swallowing. Under normal physiological conditions, continuous saliva flows, and the tongue movement mechanically and constantly removes microorganisms and renews oxygen gradients. In PD, slower swallowing and drooling most likely reduce this natural cleansing, allowing overgrowth of selected anaerobic species at the expense of overall diversity. Such conditions create microenvironments favoring taxa such as Fusobacterium, Prevotella, and Leptotrichia, which thrive in low-oxygen, nutrient-rich biofilms.

Our findings align with previous studies reporting reduced microbial diversity and increased anaerobic enrichment in saliva and oral swab samples from patients with PD [20,22]. This finding supports the assumption that reduced oral motor functions and oral clearance in PD contribute to a more anaerobic oral ecosystem. The tongue, with its irregular surface and stable biofilm, appears particularly susceptible to such an ecological imbalance, explaining why differences in terms of microbial communities and diversity between the two groups were most pronounced for tongue scrape samples.

Although the SWS flow rates did not differ significantly between groups, the saliva samples of patients with PD displayed subtle compositional changes, including enrichment of Fusobacterium and Veillonella. Since SWS samples represent newly secreted saliva, unaffected by drooling or pooling, these findings indicate that the observed dysbiosis is not merely due to low, stagnant UWS but reflects genuine alterations in the oral microbial environment. Moreover, the absence of diversity differences in SWS underscores that the dynamic, freshly produced saliva may dilute ecological differences that are more apparent on stable biofilms such as the tongue. Mastication likely had limited influence here, as the stimulated samples were obtained through standardized paraffin chewing, ensuring comparability between patients with PD and controls without PD.

Together, these findings demonstrate that both saliva and tongue scrape microbiota in patients with PD are not only enriched in the dominant genera Veillonella, Prevotella, and Streptococcus, but also harbor unique OTUs absent in the control group, particularly within the tongue scrape microbiota. This shift in community structure suggests that factors associated with PD are related to both quantitative and qualitative changes in oral bacterial populations and underline the tongue as a particularly sensitive site for detecting disease-related microbial dysbiosis.

The differential abundance analysis confirmed a consistent disease-associated enrichment of Fusobacterium species across both oral niches. The presence of F. periodonticum and F. pseudoperiodonticum in significantly higher abundance in patients with PD suggests that these taxa may play a role in shaping the altered oral microbiota composition observed in PD. Their ecological properties (strong adhesion to epithelial cells, coaggregation with other oral bacteria, and pro-inflammatory potential) make them plausible contributors to biofilm maturation and local inflammation. Notably, this enrichment in Fusobacterium-related taxa occurred despite the absence of significant differences in gingival index scores between groups. This apparent discrepancy may reflect limitations of the periodontal assessment used in this study or the influence of other, unmeasured oral inflammatory parameters not captured by the gingival index. Alternatively, given the significant difference in the plaque index observed between the groups, the enrichment of Fusobacterium species may be more closely related to increased plaque accumulation than to PD itself. While dysbiosis of the gut microbiota is increasingly implicated in α-synuclein misfolding and aggregation in PD [70–72], evidence specifically linking the genus Fusobacterium to this process is still missing. Further work is required to determine whether Fusobacterium-derived factors (e.g. lipopolysaccharides, amyloid-like proteins, or other metabolites) trigger α-synuclein misfolding in enteric neurons or facilitate its propagation to the central nervous system.

Although differences in the oral microbiota were observed between the PD group and the control group, these differences cannot be attributed with certainty to the disease itself. Alternative explanations, including differences in plaque accumulation, medication exposure, salivary dysfunction, and incompletely characterized periodontal status, cannot be excluded and may account, in part or in full, for the observed microbial differences between groups. The findings of this study should therefore be interpreted as reflecting the oral microbiota of patients with PD, rather than as evidence of a disease-specific microbial signature attributable to PD itself.

From a broader clinical perspective, the coexistence of xerostomia, drooling, impaired olfaction, and altered microbiota illustrates the complex interplay between neurological dysfunction and oral physiology in patients with PD. Reduced unstimulated salivary secretion and less frequent swallowing may contribute to both xerostomia and the accumulation of saliva anteriorly in the oral cavity, explaining the paradoxical coexistence of drooling and xerostomia often reported by patients. Cognitive decline could exacerbate these symptoms by impairing attention to oral hygiene or cue-driven swallowing. These overlapping features highlight the need for interdisciplinary management of PD, combining neurological, dental, and speech-language interventions.

Future longitudinal research should investigate whether the oral microbiota changes progress with disease severity and explore the functional relationships between microbial metabolites, salivary biomarkers, and neuroinflammatory pathways. Integration of metagenomic and metabolomic data could clarify whether specific taxa, such as Fusobacterium, contribute to systemic inflammation or neurodegeneration.

Limitations and strengths

The limitations of this study include its relatively small sample size and cross-sectional design, which preclude causal inference and reduce statistical power. Medication heterogeneity among participants with PD could also influence salivary and microbial parameters. Furthermore, the absence of a full-mouth periodontal examination is a limitation of the study, given its known influence on the oral microbiome. The study could also have benefited from the inclusion of additional oral sampling sites, such as supragingival and subgingival plaque, alongside whole stimulated saliva and tongue scraping. Plaque sampling was not feasible within the already extensive protocol, given the limited cooperation expected from this vulnerable population and the logistical challenge posed by levodopa-related on/off motor fluctuations. Nevertheless, standardized sampling conditions (collection one hour after levodopa intake, identical time windows, and consistent pre-sampling restrictions) helped minimize these effects.

The strengths of the study include a well-characterized group of patients with PD assessed with validated neurological scales (Hoehn & Yahr, UPDRS II–III), the inclusion of spouses as controls with shared environmental exposures, and detailed oral clinical assessments performed by a single calibrated examiner. The dual-site sampling approach (saliva and tongue scrape) and long-read 16S rRNA gene sequencing provided species-level resolution and robust ecological insights.

The use of SWS samples ensured comparability with previous microbiome studies and avoided the bias introduced by participants being unable to produce adequate UWS.

Conclusion

Patients with PD showed reduced unstimulated salivary flow rates, higher plaque accumulation, and distinct oral microbial profiles compared with matched controls. Furthermore, patients with PD exhibited lower diversity in their tongue microbiota and enrichment of anaerobic taxa such as Fusobacterium periodonticum and F. pseudoperiodonticum in both tongue and SWS microbiota, suggesting a shift toward a more anaerobic and biofilm-adapted ecosystem likely driven by reduced swallowing and altered salivary regulation. These findings indicate that PD is associated with both physiological and microbial changes in the oral cavity, independent of dental care adherence.

The oral environment, particularly the tongue, may thus serve as a sensitive marker of PD-related dysbiosis and a potential target for further investigation. Future interdisciplinary approaches integrating neurology, dentistry, and microbiology are needed to understand and manage the interconnected effects of motor, autonomic, and microbial dysfunction in PD.

Funding Statement

This study received funding from The Danish Parkinson Association (Parkinsonforeningen) and Helsefonden. The funding agency played no role in the submitted work.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The dataset used and/or analyzed during the current study is available from the corresponding author upon reasonable request, but a transfer of data must comply with the General Data Protection Regulation (GDPR) of the European Union (EU).

Ethical approval

The study complies with the guidelines of the Declaration of Helsinki and is in compliance with the STROBE checklist. Ethical approval was obtained from the Ethical Committee of Research Health Ethics of the Capital Region (H-20047464) and the Danish Data Protection Agency (514-0510/20-3000). The trial was registered at ClinicalTrials.gov (NCT05356845). Before inclusion, informed and written consent was obtained from the participants. The participants were also informed that no adverse events, risks, or disadvantages associated with the study were expected, and they were free to decline the offer without consequences for their treatment.