Periodontal Disease and Risk of Dementia in Medicare Patients with Hepatitis C Virus

INTRODUCTION

Alzheimer’s disease and related dementias (ADRD) refers to the major dementing illnesses. Alzheimer’s disease (AD) is the most common cause of dementia overall, occurring in 1 out of every 10 people over 65 years old. ADRD also includes Lewy body dementia, frontotemporal dementia, vascular contributions to cognitive impairment and dementia, and mixed etiology dementias [1]. As more people are entering the 65 and older age group, it is estimated that ADRD will increase in the future. These dementias are characterized by protein misfolding and propagation throughout the central nervous system (CNS). There is an increasing body of literature suggesting that chronic systemic inflammation, such as from chronic infectious states, can contribute to neuroinflammation and facilitate these neurodegenerative changes seen in dementia [2–4].

Hepatitis C virus (HCV) and periodontal disease are two common infectious states that afflict elderly adults. HCV typically targets the liver and can manifest in acute or chronic forms, the latter of which can be complicated by cirrhosis and hepatocellular carcinoma. Other organs can be involved with HCV, including the kidneys, endocrine organs, skin, and nervous system. With chronic HCV infection, 36.3% of newly diagnosed cases occur in those between 50 to 70 years of age. Most HCV infections in this population originated from exposure to HCV in blood or blood products administered in the 1960s through 1980s [5]. Periodontal disease refers to infection and inflammation of the gingiva and supporting structures of the teeth, which can lead to erosion of these structures and tooth loss [6]. It is suspected that up to 70% of adults over 65 years of age suffer from periodontal disease and has been associated with an increased risk for dementia [7–9].

There is growing literature that the infectious state, in general, has been linked to cognitive decline, with an inverse relationship between cognition and infectious burden [10, 11]. Few studies have explored whether or not multiple infectious states would be synergistic in terms of their effect on cognition, particularly among an HCV population already at increased risk for cognitive decline and dementia. The objective of this study was to examine whether periodontal disease increases the risk of developing ADRD among HCV patients. We hypothesized that the risk of developing ADRD would be significantly higher in chronic HCV patients who had an increased infectious burden with periodontal disease.

METHODS

RESULTS

We began with a cohort consisting of 615,178 FFS Medicare beneficiaries with HCV. We excluded 125,142 patients who were not continuously enrolled in Part A and Part B. We then excluded 36,724 individuals who were diagnosed with ADRD before the index date and 841 patients who did not seek any periodontitis care in the year prior to the index year (i.e., those who did not have chronic periodontitis). Because of missing ZIP-level variables, including rurality, income, and education, 12,711 individuals were excluded. A total of 439,760 Medicare beneficiaries met our study inclusion criteria.

Table 1 presents characteristics of the study sample. Approximately 38% were female, 22% were African-American, and 44% were aged 65 and older. The most common comorbidities were hypertension (59%), drug and alcohol related disorder (44%), cardiovascular diseases (35%), and tobacco use disorder (34%). Patients with chronic periodontitis were more likely to be white or Hispanic, female, and eligible for both Medicare and Medicaid, live in areas with high income and a high percentage of college graduates, and have comorbid conditions, compared with those without chronic periodontitis. The incidence rate of ADRD in the periodontal disease group was higher than in patients without periodontal disease (10.84% versus 9.26%, p < 0.001, and those with periodontal disease earlier compared to those without periodontal disease (13.99 versus 21.60 months, p < 0.001).

Figure 1 shows the Kaplan-Meier curve comparing the unadjusted rate of ADRD incidence between patients with and without chronic periodontitis. Chronic periodontitis was associated with a higher proportion of incident ADRD compared to those without chronic periodontitis.

OPEN IN VIEWER

Fig. 1

Kaplan Meier Curve of Alzheimer’s Disease and Related Dementia Incidence.

The results from Cox regression are shown in Table 2. The unadjusted hazard of developing ADRD from Model 1 was 78% higher in patients with chronic periodontitis compared with those without chronic periodontitis (unadjusted hazard ratio (HR) = 1.78, 95% CI: 1.72–1.84). When adjusted for patients’ demographic and neighborhood characteristics (Model 2), the hazard of ADRD incidence in individuals diagnosed with chronic periodontitis was 70% higher than in patients without chronic periodontitis (adjusted HR = 1.70, 95% CI: 1.65–1.76). In the fully adjusted model (Model 3), patients with chronic periodontitis had 35% higher hazard of developing ADRD than those without chronic periodontitis (adjusted HR = 1.35, 95% CI: 1.30–1.40).

Several patient factors were significantly associated with ADRD incidence. In the fully adjusted model (Model 3), the hazard of developing ADRD increased with every 5-year increase in age. Notably, compared with beneficiaries younger than age 55, the hazard of ADRD incidence was 6 times higher in those aged 75 and older. Females and patients residing in neighborhoods with higher income or rural areas had lower hazard of developing ADRD. In contrast, being African-Americans, eligible for both Medicare and Medicaid, and living in areas with higher education increase the hazard of ADRD incidence. The presence of most comorbidities, such as hypertension, diabetes, cardiovascular diseases, and drug and alcohol disorder, were associated with higher hazards of developing ADRD, while patients with tobacco use disorder had lower hazard of developing ADRD.

DISCUSSION

This study is, to our knowledge, one of the first analyses that has examined the relationship between periodontal disease exposure to incident ADRD among an HCV population. We found that the risk of developing ADRD was significantly higher in the HCV population with exposure to periodontal disease. This increased risk for ADRD in those exposed to periodontal disease was noted through all three models, highest in the univariate model and reduced but still positive in the fully adjusted model accounting for cardiovascular risk factors (Table 2). Although this increased risk was seen through all age subgroups, the subgroup within the mid-life age range (i.e., 50–64) carried the highest risk (Table 4). A potential reason for this finding of increased risk for ADRD in the mid-life range age group may be that this population is sicker since acceptance into Medicare prior to 65 years must include the presence of significant disabilities or certain medical conditions such as end-stage renal disease or amyotrophic lateral sclerosis.

The association between periodontal disease and ADRD remained significant for the youngest age group of individuals in midlife after adjusting for demographic and medical covariates. This may potentially have clinical implications, possibly suggesting that primary prevention of dementia through modifying risk factors may need to begin during midlife or even earlier [13, 14]. There is a paucity of data regarding whether treating periodontal disease in midlife or earlier may decrease risk of ADRD, particularly among an HCV population who are already at increased risk for dementia. As previously mentioned, periodontal disease is caused by a dysbiosis of the commensal oral flora, and there are a number of studies exploring the role of probiotics in improving outcomes in hepatitis patients in general, although most of these studies have sought to examine liver chemistries and hepatic encephalopathy, rather than incident ADRD [15, 16]. Few studies have examined the effect of probiotic supplementation on cognitive function among individuals with AD and there is insufficient evidence at the present time to support the use of probiotics for individuals with ADRD [17].

Our results are consistent with previous studies that have found positive associations between infectious burden from multiple infections and poor cognition [2–4]. Our findings suggest a potential synergistic effect between periodontal disease and HCV on risk for dementia after up to 4 years of follow-up, with patients who have periodontal disease being diagnosed with ADRD earlier than those without it. Several mechanisms may explain this association. Systemic infections result in a pro-inflammatory state, which can lead to loss of integrity of the blood-brain barrier, with resultant penetration of the CNS by inflammatory cells and microglial activation. It is possible that polymicrobial infections enhance this proinflammatory cascade and prolong its effects [18]. This pro-inflammatory milieu appears to be facilitated through a combination of systemic introduction of the periodontal bacteria as well as alterations in the immune system and host metabolism that sustain the inflammatory response [19]. Periodontal disease can worsen liver disease through the generation of inflammatory cytokines and reactive oxygen species by the periodontal tissue in response to the pathogenic bacteria. Chronic, systemic inflammation also leads to insulin resistance, which can lead to accelerated atherosclerotic disease and dementia [19–21].

A notable strength of our study is the large sample size in the national claims dataset. We were able to adjust for a set of covariates in the full models, including demographic variables and vascular risk factors. Even after adjusting for all those covariates, a significant association was still found between periodontal disease and ADRD. Furthermore, we performed stratified analyses by age group.

One limitation of our study is potential misclassification since the diagnoses of ADRD and periodontal disease are based on billing codes. The actual onset of symptoms of ADRD is likely to be earlier than the diagnosis date in the claims data; however, this is unlikely to be differential based on periodontal disease status (i.e., unlikely to be a differential bias and impact results). Furthermore, the clinical severity of the dementia and clinician’s decision-making as to the diagnoses are not known. The temporal onset of periodontal disease could not be exactly established; however, this was mitigated in part by clarifying the chronicity of the ICD codes (i.e. chronic periodontal disease or history of periodontal disease). Further studies are needed to clarify the effect of duration of periodontal disease on risk for dementia.

Our study suggests a potential synergistic effect of periodontal disease and HCV on risk for ADRD. It is possible that periodontal disease may be a modifiable midlife risk factor among an HCV population at risk for dementia. This possibility warrants further investigation in future studies.