Is There an Association Between Parkinson’s Disease and Periodontitis? A Systematic Review and Meta-Analysis

INTRODUCTION

Parkinson’s disease (PD) is a progressive neurodegenerative disorder manifested by bradykinesia, resting tremor, rigidity, and postural instability, affecting millions of elderly adults globally. Between 1990 and 2016, the age-standardized prevalence, disability-adjusted life-years rates and mortality of PD have skyrocketed in almost every region worldwide, increasing the heavy global financial burden [1, 2]. Despite the specific pathogenesis remaining elusive, inflammatory and immunological factors have been suggested to participate in its underlying mechanisms [3, 4].

Periodontitis is a chronic inflammatory disease of the oral cavity that is triggered by the formation of dental plaque biofilm [5]. Advanced periodontitis is a leading cause of loss of teeth and supporting structures, affecting approximately 10–15% of the global population [6]. According to the data from the 2009–2012 National Health and Nutrition Examination Survey, its overall prevalence among Americans over 30 years was estimated to be as high as 46% [7, 8]. It has been observed that the prevalence of total periodontitis rises from the 30–34 age group to the aged 65+ age group [7]. The progression of age, coupled with reduced dexterity, results in the accumulation of dental plaque, leading to more frequent and severe periodontal inflammation [9], which may elevate the longitudinal risk of periodontitis. Moreover, periodontitis has been reported to be correlated with several systemic diseases, such as diabetes mellitus, cardiovascular diseases [10], Alzheimer’s disease (AD) [11], and cancer [12].

Various articles have reported the high prevalence of periodontitis and poor oral health in individuals with PD. For instance, a nested case-control study reported that the prevalence of periodontal diseases attained 48.73% in a cohort of 4765 PD patients [13]. In a cross-sectional case-control study, periodontitis was detected in 98 out of 104 PD patients [14]. Several factors could potentially explain the link between the prevalence or risk of periodontitis and PD. As is well documented, polypharmacy, cognitive impairment, motor dysfunction, and the subsequent deducing dental care elevate the risk of dental caries and periodontal disorders [15]. Sialorrhea, a commonly observed symptom of PD, is also hypothesized to promote periodontitis and other oral diseases, attributable to hypersalivation altering the composition of oral microbiota (like K. oralis) in PD cases and facilitating local inflammation in the oral cavity [16]. Additionally, the prevalence of sialorrhea has been reported to increase with age and disease duration in PD patients [17], insinuating that it may mediate the impact of age and PD duration on periodontitis risk. Similarly, gender is also considered a key factor in the development of sialorrhea, thereby influencing periodontitis risk, which can be supported by the finding that males with PD were twice more likely to develop sialorrhea compared to females [18]. Hyposmia, characterized by a reduced sense of smell in PD patients [19], which deteriorates significantly with age [20] and is often accompanied by taste dysfunction, contributes to prolonged mealtimes and inadequate chewing [15], thus the increasing risk of periodontal diseases. Moreover, since female gender might be a contributing factor for dysphagia [15, 21], another plausible explanation for the effect of gender on periodontitis risk is that dysphagia may result in acid reflux, which could trigger perimylolysis tooth erosion and potentially increase the periodontitis risk [22]. In addition, Fukayo et al. [23] reported that female PD patients brushed their teeth more frequently than male patients and tended to have a lower decayed, missing and filled teeth (DMFT) index than the male group, indicating that the alterations in dental care behavior may also be a reason for gender influencing periodontitis risk.

Despite numerous articles investigating the association between PD and periodontitis, their findings have been contradictory. While certain large-sample observational studies have noted that periodontal inflammatory disease is correlated with an elevated risk of developing PD [24, 25], other studies have yielded conflicting results. For example, a Mendelian randomization (MR) study revealed no convincing genetic association between the two diseases [26]. Meanwhile, the bidirectional relationship between PD and periodontitis has not been systematically assessed.

Therefore, we sought to systematically review studies focusing on PD and periodontitis and perform a meta-analysis to comprehensively investigate the bidirectional association between the two diseases and summarize the oral status of PD patients.

METHODS

RESULTS

Literature search

The literature search yielded 2,708 studies, and three additional publications were identified by reviewing reference lists. After excluding duplicates, a total of 1,985 unique titles and abstracts were screened. Subsequently, the full text of 52 potentially eligible articles was reviewed. Following a thorough assessment, 18 studies were excluded: four studies using the same data from another publication, four studies reporting dementia or cognitive impairment as the outcome, and ten conference abstracts did not report quantitative data. Therefore, 34 studies were included in the systematic review. Concerning studies with data on control groups, a quantitative meta-analysis was performed on 24 studies (involving 7,720,793 participants) reporting the same outcomes. The process of study identification is illustrated in the PRISMA flow chart [29] in Fig. 1.

Fig. 1

PRISMA flow diagram delineating study identification.

Association between periodontitis and risk of PD

The overall HR for the 3 cohort studies (involving 7,298,955 participants) was 1.13 (95% CI 0.88–1.45, p = 0.3436, I² = 96%; Fig. 2A), demonstrating substantial evidence that periodontitis is not associated with the incidence of PD. In line the results of cohort studies, the combined OR, including 4,844 participants from six retrospective studies and one Mendelian randomization study [26] also showed an overall null effect (OR 1.94, 95% CI 0.55–6.90, p = 0.3040, I² = 81%; Fig. 2B) for the association between periodontitis and the risk of PD.

Fig. 2

Forest plots of meta-analysis for the association between periodontitis and risk of Parkinson’s disease, combining the HR of three cohort studies (A), and combining the OR of six retrospective studies (B).

Association between PD and prevalence of periodontitis

Owing to the diverse and heterogeneous outcomes in the included studies, periodontal clinical parameter values were integrated to evaluate the prevalence of periodontitis and the periodontal status of PD patients. The pooled PPD in PD patients was significantly higher than that in control groups (SMD = 1.10, 95% CI 0.53–1.67, p = 0.0002, I² = 85%; Fig. 3A). Similarly, the CAL of PD patients was significantly higher (SMD = 1.40, 95% CI 0.55–2.26, p = 0.0013, I² = 92%; Fig. 3B) compared to control groups, and the PI of PD patients was also significantly higher (SMD = 0.81, 95% CI 0.22–1.39, p = 0.0072, I² = 87%; Fig. 3C). Besides, although the BoP % was higher in PD groups than that in control groups, there was high heterogeneity and no statistically significant difference (SMD = 0.80, 95% CI –0.65–2.25, p = 0.2803, I² = 96%; Fig. 3D). Above all, PD patients exhibited higher values of periodontal clinical parameters, indicating a higher prevalence of periodontitis among PD cases.

Fig. 3

Forest plots of meta-analysis for PPD (A), CAL (B), PI (C), and BoP% (D) in patients with Parkinson’s disease.

Sensitivity analysis

The leave-one-out sensitivity analysis revealed that no single study altered the pooled effect of HR and OR (Fig. 4A, B), proving the robustness of our results, which demonstrated no significant association between periodontitis and the risk of PD. Additional results of the sensitivity analyses on other outcomes (PPD, CAL, PI, and BoP%) are provided in the Supplementary Material.

Fig. 4

Leave-one-out sensitivity analysis of studies focusing on periodontitis and the risk of Parkinson’s disease.

Publication bias assessment

Due to the limited number of included studies for each outcome in the meta-analysis, a contour-enhanced funnel plot was only performed for the seven retrospective studies regarding periodontitis and the risk of PD. Asymmetry of the funnel plot was observed. In the contour-enhanced funnel plot (Fig. 5), the hollow dots filled by trim-and-fill approach depicted that one filled study was situated in sky-blue areas indicating high statistical significance (p < 0.01), whereas another filled study was located in white areas of no statistical significance. This discrepancy demonstrates that the potential publication bias may stem from the study conducted by John et al. [30]. However, even after excluding this study in the leave-one-out sensitivity analysis, the combined OR remained statistically insignificant (OR = 1.94, 95% CI 0.47–8.10, p = 0.3614; Fig. 4B), which further supports the stability of our results.

Fig. 5

Contour-enhanced funnel plots assessing the potential publication bias on studies concerning periodontitis and the risk of Parkinson’s disease.

DISCUSSION

The principal findings of our meta-analysis imply that patients with periodontitis do not tend to have a higher risk of developing PD, and that PD patients might not confer a higher risk of periodontitis. Nonetheless, it was observed that PD patients presented a higher prevalence of periodontitis and other periodontal diseases, as well as poorer oral hygiene.

Our findings, which demonstrate no association between periodontitis and PD risk, are in agreement with a prospective cohort study recruiting 15,528 patients [31]. This study illustrated that the risk of PD was not significantly associated with dental plaque whether in males (pooled HR = 1.23, 95% CI 0.81–1.86, p = 0.3309, I² = 0%) or in females (pooled HR = 0.86, 95% CI 0.56–1.31, p = 0.4788, I² = 0%). However, due to the heterogeneity of the outcomes (the presence of dental plaque but not confirmed periodontitis), this study was only included for the systematic review but not in the quantitative meta-analysis.

Previous studies have stressed the significant role of periodontitis in the pathogenesis of PD, demonstrating that periodontitis triggers the systemic translocation of bacteremia and inflammatory mediators, resulting in the breakdown of the blood-brain barrier and activation of microglia, eventually driving in necrosis and apoptosis of dopaminergic neurons and contributing to the mechanism of PD [32, 33].

In contrast to these viewpoints, we posit that the subsequent inflammatory processes may play a dominant role in the progression of PD, rather than periodontitis itself, with periodontitis serving as one of the numerous initiating factors of inflammation. Our hypotheses are supported by the following evidence.

As was reported in one of the included studies [16], although there were no differences in dental or periodontal parameters between 20 PD patients and 20 healthy controls, the elevated pro-inflammatory cytokine levels, such as interleukine-1β and tumor necrosis factor-α, were still observed in the gingival crevicular fluid of PD cases compared with controls. This reflected the presence of active peripheral inflammatory responses in the oral cavity of PD patients, irrespective of the presence of periodontitis, thus partially corroborating our assumptions.

Furthermore, a study [34] enrolling 37 PD patients reported an increase in the number of various inflammatory cells, including leukocytes, basophils, and segmented neutrophils, in PD cases complicated by periodontitis. Intriguingly, another study involving 51 PD participants described the levels of circulating high-sensitive C-reactive protein were increased in the periodontitis group, whereas the lymphocyte levels were elevated in the group without periodontitis [35]. However, these studies did not include a periodontitis control group without PD, which prevented the investigation of whether the rising levels of inflammatory cells and parameters were primarily attributed to PD or periodontitis itself. Thus, further studies are required to clarify the mediating role played by inflammatory factors in these two diseases.

Apart from that, previous research has emphasized the essential role of inflammation in the development of PD. It has been suggested that systemic inflammation events could induce neuronal death and contribute to the progression of neurodegenerative diseases [36]. Specifically, activated glial and peripheral immune cells could mediate neuroinflammatory processes which are deleterious to neurodegeneration and stimulates the death of dopaminergic cells [3]. Besides, inflammation appears to be the bridges connecting several chronic diseases with PD. For example, there are shared inflammation dysregulated pathways linking diabetes with PD [37]. Additionally, chronic kidney disease promotes PD development by exacerbating inflammation and oxidative stress via excessive activation of immune response caused by metabolic toxins [38]. Hence, further studies should be undertaken to investigate the extent to which inflammation plays a role in the progression of PD.

Intriguingly, it is worthwhile noting that our results regarding causal inference between PD and the periodontitis risk were primarily inferred from the MR study, since it was the sole study in the literature examining their association. However, MR study could only assist in confirming their weak association with genetic liability. As documented in the MR study, MR may be powerless in the presence of a strong environmental interaction and may not be able to elucidate the influence of a disease on the occurrence of another disease during a specific period of lifetime, thereby resulting in potentially inaccurate conclusions [26]. In line with this perspective, the result of the lack of association between PD and the risk of periodontitis should be interpreted with caution. Considering the higher prevalence of periodontitis and worse oral cavity health among PD patients, we speculate that it is the several environmental factors that mediate the spurious weak association between PD and periodontitis risk, and that these factors cannot be evaluated at the genetic level using an MR study. For instance, motor deficits in PD patients might be an influencing environmental factor. Indeed, the manifestation of motor impairment, hypokinesia, and involuntary movements in PD patients hinder their capability to maintain their daily oral hygiene, ultimately giving rise to poor oral health prognosis and deterioration in their quality of life [39]. Meanwhile, insufficient dental management induced by cognitive impairment may also be an environmental factor placing PD patients at risk for dental caries and periodontal diseases [15], as neglect of oral cavity disorders by PD cases, their caregivers and their professional physician has been reported [40]. Furthermore, other common symptoms of PD patients, like hyposmia, taste dysfunction [19], dysphagia [21], and masticatory disturbance [41], would contribute to longer meal times and more inadequate mastication, inducing a reduction in food intake quantity and quality, thereby causing nutritional imbalance and deterioration of oral health disorders in PD [15]. Last but not least, fluctuations in eating habits, such as preference for soft sticky foods [42] or sweets [43], would also conduce to the formation of cariogenic environments and subsequently lead to the occurrence of periodontal diseases [15]. In a nutshell, the aforementioned non-negligible environmental factors may potentially account for the high prevalence of periodontal diseases and poor oral health of PD patients, and more studies are necessitated to explore their effect.

The association between the risk of periodontitis and another degenerative disease, AD, has yielded inconsistent conclusions as well. In addition to a longitudinal study determining no statistical differences in several periodontal parameters between AD patients and healthy controls during 3 years of follow-up [44], no significant association between AD and periodontitis risk (RR = 1.4, 95% CI 0.9–2.1) was also reported by another study including 174 AD patients [45]. However, a population-based cohort study recruiting 8,640 individuals demonstrated an inverse association between AD and periodontitis risk (HR = 1.667, 95% CI 1.244–2.232, p = 0.0004) [46]. Hence, the relationship between AD and the risk of periodontitis also remains paradoxical and requires further investigation. Likewise, since our result on the weak association between PD and periodontitis risk may be biased and unreliable in our study, their casual association remains to be further investigated and clarified by additional well-designed and large-scale observational studies.

Our study possesses several strengths. To the best of our knowledge, this is the first systematic review and meta-analysis to comprehensively evaluate the bidirectional association between periodontitis and PD. Moreover, integrating the evidence from a Mendelian randomization study enhances the robustness and persuasiveness of our findings, since MR is widely recognized as providing higher-quality evidence compared to conventional observational studies by utilizing genetic variation as an instrumental variable to detect causality. Furthermore, contrary to previous theories, our intriguing findings provide no compelling evidence to support a bidirectional correlation between the incidence of the two diseases, neither in clinical manifestation nor in genetic liability [26], which may challenge traditional hypotheses concerning their potential relationship. Lastly, our study demonstrated that PD patients indeed suffer from poor oral hygiene, which has implications for guiding the management of dental diseases and improving oral care among PD patients.

Some potential limitations of our study should also be acknowledged. First of all, high heterogeneity was detected in our results, which can be chiefly attributed to variations in diagnostic criteria for periodontitis and measurements of outcomes. The lack of consensus on the definitions of periodontitis, discrepancies in periodontal examination methods, and the heterogeneity in dental conditions pose challenges in estimating the actual prevalence of periodontitis [47] while these factors were also an important source of high heterogeneity in our results and the asymmetry of the funnel plot. Thus, caution should be exercised in interpreting our results due to the existing heterogeneity. However, sensitivity analysis and contour-enhanced funnel plots supported the stability of our findings, reinforcing the credibility of our results. Secondly, as previously mentioned, the weak association between PD and the risk of periodontitis may be unreliable and should be interpreted with caution. This is primarily due to only one MR study addressing this topic while large-scale cohort studies focusing on the risk of periodontitis in PD patients with long-term follow-up are scarce, which also prevented an accurate reviewing and assessment of the incidence of periodontitis in PD cases. Finally, despite the limited number of included studies, grey literature databases comprising unpublished articles were not searched, given the uncertainty in their level of evidence quality. Despite these limitations, our study comprehensively evaluated the association between PD and periodontitis and provide profiles of the oral status of PD patients, which could aid dentists and specialist physicians in formulating considerate dental care strategies for PD patients to optimize their overall quality of life.