Association of Gum Treatment with Cognitive Decline and Dementia Risk among Older Adults with Periodontal Symptoms: A 12-Year Prospective Cohort Study

Dementia is one of the major causes of disability and mortality among older people [1]. As of 2022, over 55 million people are living with dementia worldwide, and the number is estimated to approach 152 million by 2050 [2]. Although dementia prevention has primarily focused on controlling modifiable cardiometabolic, dietary, and other lifestyle factors [1], a growing body of literature suggests that oral health conditions such as tooth loss [3‒5], periodontal disease [6, 7], and poor self-reported oral health [8] are also associated with accelerated cognitive decline and increased risks of dementia.

Periodontal disease, characterized by chronic gum inflammation, is especially prevalent in older adults in the USA, affecting 68% of those aged ≥65 [9]. It not only is a local oral health issue but also has systemic implications [10]. For example, periodontal pathogens and the resulting host immune response led to increases in pro-inflammatory cytokines such as interleukin-1β, interleukin-6, and tumor necrosis factor-α [11]. These cytokines may compromise the integrity of the blood-brain barrier and promote neuroinflammation [12]. Additionally, systemic inflammation resulting from periodontitis has been associated with surrogate markers of neurodegeneration, such as decreased brain volume, cortical thinning, and white matter hyperintensities [12, 13]. Indeed, periodontal diseases, cognitive decline, and dementia share risk factors, including cardiovascular disease, diabetes, obesity, smoking [14], as well as stress, depression, and inflammation [15]. Gum treatment aims primarily to reduce pocket depths and the total subgingival biofilm, encompassing various periodontal procedures, ranging from nonsurgical methods like deep cleaning and antibiotics to surgical interventions (flap surgery and tissue or bone grafts) [16]. However, an increasing number of studies indicated that gum treatment leads to apparent decreases in the circulating concentrations of inflammatory markers and considerable improvement in vascular endothelial function [17‒20]. Despite growing evidence of the untreated periodontal disease-dementia link, there is a paucity of studies specifically examining the association of gum treatment with cognitive outcomes.

The association between periodontal disease and cognitive outcomes may vary across sociodemographic subgroups. There is some evidence, for example, that sex moderates this association, with the relation between periodontitis and dementia risk stronger among older women than men [21]. A nationwide retrospective cohort study indicated that periodontitis was a stronger predictor of dementia at relatively older ages than relatively younger ages [22]. In addition, there are racial/ethnic differences in the rate of cognitive decline and dementia risk in later life, as well as differences by educational attainment [23]. When identifying risk factors for cognitive decline and dementia, it is critical to determine whether the level of the risk is similar across groups with more and less vulnerability.

This study aimed to examine the association of gum treatment with cognitive decline and dementia risk among older adults in the USA. Additionally, we address whether these associations vary by age, sex, race, ethnicity, education, income, and the severity of periodontal symptoms.

Data were drawn from the Health and Retirement Study (HRS). Starting in 1992, HRS is a biennial longitudinal survey of a nationally representative sample of community-dwelling adults aged 50 years and older in the USA [24], and it oversamples black and Hispanic older adults. HRS collects detailed information on demographics, economic, work, family, health behaviors, and health conditions every 2 years. The HRS has high response rates (82–90%) in each wave [24]. In 2008, the HRS administered a “Dental Health, Access to Care, and Utilization Module” (Dental Health Module hereafter) to a randomly selected subsample of 1,267 individuals to collect extensive information beyond what was available in the HRS Core survey, encompassing aspects like dental condition, oral health-related quality of life, dental care utilization, and the specific types of treatments received. The HRS was approved by the Institutional Review Board at the University of Michigan. A detailed description of HRS has been published elsewhere [25].

Participants were selected for analysis if they (1) completed the Dental Health Module in 2008 (baseline); (2) aged ≥50; (3) reported being affected by periodontal conditions; (4) did not have dementia at baseline; (5) had at least one wave follow-up through the 2020 wave. A total of 215 respondents were excluded from our main analysis because they did not report any periodontal symptoms. These participants without periodontal symptoms were relatively younger, more likely to be female, and had higher income and education levels, fewer health conditions, and higher cognitive function scores compared to those with periodontal symptoms (no periodontal symptoms: mean [standard deviation, SD], 18.32 [5.44] vs. had periodontal symptoms: mean [SD], 16.51 [3.87]). Overall, a total of 401 participants were finally excluded based on all five criteria. The excluded participants had fewer years of education, less income, and poorer health conditions than the 866 participants included in the analysis. There were no differences in age, sex, or ethnicity between those two groups (p > 0.05; see online suppl. Table S1; for all online suppl. material, see https://doi.org/10.1159/000540086). A flowchart illustrates a sample section process (Fig. 1).

Previous literature indicated that self-reported sensitive teeth and avoiding eating particular food were associated with mild to moderate periodontal disease, while bleeding teeth indicated more severe periodontal disease [26, 27]. Periodontal symptoms in this study were therefore determined by an index of three questions regarding their oral and dental conditions during the last year [28, 29]. Specifically, participants were asked how often (1) your gums had bled when you brushed your teeth, (2) you had avoided particular foods because of problems with their teeth or mouth, (3) your teeth or gums had felt sensitive to hot, cold, or sweets, on a scale range from 1 to 5 (very often to never). Items were reverse-scored in the direction of more severe periodontal symptoms. The summary scale ranges from 3 to 15, with higher scores representing worse periodontal health (α = 0.81). Based on previous studies and the distribution of the responses [28, 29], we classified participants that scored ≥10 as “affected by periodontal symptoms,” and participants scored <10 as “not affected by periodontal symptoms.”

Information on gum treatment was collected using the following question: “since your last interview or in the last 2 years, have you seen a dentist or other dental provider (including periodontists, oral surgeons, prosthodontists, dentists, etc.) for gum treatment?” Based on their responses, participants were classified into two groups: 1 = had gum treatment or 0 = no gum treatment.

We utilized the researchers-contributed dataset that includes cognitive scores and dementia status and is publicly accessible on the HRS website [30]. Cognitive function was assessed using the Telephone Interview for Cognitive Status (TICS), a tool comprising three sections: immediate and delayed recall (20 points), serial 7 subtraction (5 points), and backward counting (2 points). The TICS summary scores, with a maximum of 27 points, were computed by adding the individual scores from each test. Higher scores reflect better cognitive functioning. Previous research has indicated the efficacy of TICS in dementia screening [31, 32]. Dementia statuses were identified using the Langa-Weir algorithm, categorizing individuals with TICS scores ≤6 points as having dementia [31, 32].

For participants unable to respond themselves, proxy respondents were used. Their cognitive function was evaluated through assessments of five instrumental daily activities (taking medication, cooking, using the telephone, managing money, and shopping; range, 0–5 points), memory levels (0 = excellent, 1 = very good, 2 = good, 3 = fair, 4 = poor; range, 0–4 points), and the interviewer’s assessment of the respondent’s difficulty completing the interview because of cognitive limitations (0 = none, 1 = some, and 2 = prevents completion; range, 0–2 points). The total proxy assessment scores range from 0 to 11, with higher scores indicating lower cognitive abilities. Participants with proxy scores ≥6 were classified as having dementia [31, 32].

Covariates were selected according to prior research on the topic [3, 33]. Participants’ sociodemographic characteristics included age (in years), sex (men/women), race (white, black, and other), ethnicity (Hispanic/other), education (years of schooling), and annual household income. Health-related behaviors included current smokers (yes/no) and social isolation. Each participant received a social isolation score (0–5) based on living arrangement, marital status, contact frequency with children, family members, friends, and involvement in social groups. Those in the top quintile (scoring ≥2) were classified as socially isolated [34, 35]. Health conditions included obesity (yes/no), hypertension (yes/no), diabetes (yes/no), and depressive symptoms measured by the Center for Epidemiologic Studies Depression Scale-8 [36]. Functional status was measured by disability in activities of daily living (yes/no). Dental insurance coverage was assessed by affirming the question: “do you have any insurance that covers any part of dental care?” The detailed definition of each covariate is shown in online supplementary Table S2.

All analyses were conducted using Stata MP 17.0 (StataCorp LLC., College Station, TX), with a two-tailed value of p < 0.05 considered statistically significant. Characteristics of participants who had/no gum treatment were compared using two-sample t tests for continuous variables and the χ² tests for categorical variables. Kaplan-Meier plots were used to demonstrate the cumulative incidence rate of dementia with the log-rank test to compare the probabilities of dementia between the two groups. Participants were censored at the date of developing dementia, death, or the end of follow-up (2020 wave), whichever came first.

Associations between gum treatment and longitudinal cognitive decline were estimated with linear mixed-effects models in which a participant-specific random intercept and slope of gum treatment were applied. The linear mixed-effects models were constructed with gum treatment, time, and a product term of gum treatment × time as predictive factors and cognitive scores as the dependent variable. Participants with no gum treatment defined the reference group (model 1). Model 2 was adjusted for sociodemographic factors (age, sex, race, ethnicity, education, and income). Model 3 was additionally adjusted for behaviors and health conditions (smoking, social isolation, obesity, depressive symptoms, hypertension, diabetes, activities of daily living disability, and periodontal symptoms). Model 4 was additionally adjusted for dental insurance coverage. To account for potential differences between self-reported and proxy-reported cognitive assessments, we included a binary indicator variable for proxy-reported status (1 = proxy-reported, 0 = self-reported) as a covariate in all models.

Cox proportional hazard models were used to estimate hazard ratios (HRs) for dementia risks over a 12-year follow-up. The proportional hazard assumption was tested using the Schoenfeld residual test and complementary log-log plots. Gum treatment was entered as the predictor of incident dementia over a 12-year follow-up, with participants with no gum treatment being treated as a reference (model 1). The Cox models were adjusted for the same covariates with the same sequence as linear mixed-effects models.

We tested whether the associations of gum treatment with cognitive decline and dementia were moderated by age, sex, race, ethnicity, education, income, or the severity of periodontal symptoms. Interactions were tested separately for each potential moderator, and all other covariates were included in the model (e.g., when the interaction between gum treatment and age was tested, other sociodemographic factors, behavioral factors, and health conditions were included as covariates).

The study involved 866 participants with an average age of 67.7 (SD 10.1) years. Of these, 314 (36.3%) had gum treatment, while 552 (63.7%) had not (Table 1). Participants with gum treatment tended to have higher education and income levels, more dental insurance, and better periodontal health (p < 0.001), with no significant differences in age, sex, race, ethnicity, smoking habits, social isolation, health conditions, or baseline cognitive function (p > 0.05). The gum treatment group had a dementia incidence rate of 7.39 per 1,000 person-years, compared to 12.94 for those without treatment over a median follow-up of 9 (interquartile range, 6–10) years. Those receiving gum treatment showed a significantly lower dementia incidence (Fig. 2; log-rank test; p = 0.009).

Table 2 presents the results from the linear mixed-effects regressions. As shown in unadjusted model 1, gum treatment was not significantly associated with cognitive function at baseline (β coefficient, 0.260; 95% CI, −0.364 to 0.883; p = 0.453). However, compared with participants who did not have gum treatment, those with gum treatment experienced a slower decline in cognitive function (β coefficient, 0.041; 95% CI, 0.021 to 0.061; p < 0.001). In the final model 4 that adjusted for dental insurance coverage, we found that participants with gum treatment experienced a decline in TICS score that was on average 0.025 points less per year (95% CI, 0.005 to 0.044; p = 0.022) than those without gum treatment.

Table 3 shows the results of the Cox proportional hazard model. Compared with the participants without gum treatment, the HRs for dementia were 0.56 (95% CI, 0.36–0.88; p = 0.008) for those with gum treatment. The association was independent of sociodemographic factors (model 2), behavior and health conditions (model 3). In the final model (model 4) adjusting for dental insurance coverage, the association was attenuated but remained significant (HR, 0.62; 95% CI, 0.41–0.93; p = 0.012).

There was little evidence that the association between gum treatment and cognitive decline varied across sociodemographic groups and periodontal symptoms. The association between gum treatment and cognitive decline was not moderated by age (p = 0.079 for interaction), sex (p = 0.617 for interaction), race (black vs. white; p = 0.341 for interaction), ethnicity (Hispanic vs. other; p = 0.097 for interaction), education (p = 0.106 for interaction), income (p = 0.082 for interaction), or the severity of periodontal symptoms (p = 0.221 for interaction). Additionally, the association between gum treatment and dementia risk was not moderated by age (p = 0.484 for interaction), sex (p = 0.207 for interaction), race (black vs. white; p = 0.527 for interaction), ethnicity (Hispanic vs. other; p = 0.844 for interaction), education (p = 0.265 for interaction), or the severity of periodontal symptoms (p = 0.924 for interaction). However, we found a significant interaction between income and gum treatment (p = 0.011 for interaction) on dementia risk.

Using data from a 12-year prospective cohort, we found that compared to older adults who did not receive gum treatment, those with gum treatment experienced a decline in TICS score that was on average 0.025 points less per year and a 38% lower incidence of dementia. Of note, these associations persist irrespective of common dementia risk factors and dental insurance coverage. Furthermore, these associations were consistent across participants with different severity of periodontal symptoms and sociodemographic characteristics except for the income level.

In our study, only 36.3% of participants with periodontal symptoms reported receiving gum treatment. While this proportion may seem low, it is consistent with previous research on dental care utilization and periodontal treatment among older adults in the USA [9, 37, 38]. The low rates of gum treatment may be attributed to factors such as limited access to dental care, lack of dental insurance, and low awareness of the importance of oral health [39, 40].

Periodontal pathogens and the host immune response increase pro-inflammatory cytokines such as interleukin-1, interleukin-6, and tumor necrosis factor-α [41]. The pro-inflammatory cytokines might compromise the blood-brain barrier and cause inflammatory reactions in the central nervous system [17]. Therefore, the possible mechanism between periodontal disease and cognitive outcomes could be the linkage of inflammation and pro-inflammatory cytokines [42]. Additionally, periodontal disease has previously been associated with cardiometabolic risk factors that are also implicated in cognitive decline and dementia. For example, periodontal disease is associated with a higher risk of raised blood pressure, high blood glucose, abnormal lipids, and obesity [43, 44]. However, an increasing number of studies indicated that gum treatment leads to apparent decreases in the circulating concentrations of inflammatory markers and considerable improvement in vascular endothelial function [17‒20]. For instance, anti-infective periodontal treatment can result in short-term modest reductions in systemic inflammation [45], and multicomponent periodontal treatments significantly reduce hemoglobin A1c, other than probing depth and clinical attachment level [46].

The protective effect of dental treatment on dementia risk was investigated in a recent retrospective case-control study [47]. Chen et al. [47] identified 18,930 patients with a dementia-related diagnosis from the Taiwan National Health Insurance Research Database and found regular scaling can reduce the incidence of dementia in those aged 40 years and older. Possible explanations for the link between gum treatment and cognitive health are that periodontal treatment reduces cardiometabolic risk factors and systemic inflammation markers among older adults with periodontal conditions, which are strongly connected to cognitive decline and dementia risk. Second, the impact of periodontitis on nutritional intake and diet patterns should be considered [48]. Periodontitis often leads to tooth loss, impairing chewing function. Effective periodontal treatment can prevent tooth loss, thereby maintaining good chewing function and benefit from a diet rich in fiber, which is known to enhance cognitive performance [49]. Third, older adults with poor oral health are less engaged in social activities, such as church and family gathers [50]. Studies have found that older adults who are socially isolated or feel lonely may be more susceptible to developing cognitive impairment or dementia [51, 52]. Finally, the costs of gum treatment are considerable for individuals affected with periodontal diseases. Older adults who had gum treatment are more likely to be socioeconomically advantaged, as demonstrated in our study (Table 1). Psychological or occupational stress of socioeconomically disadvantaged older adults leads to accelerated cognitive decline and dementia [53]. The cumulative effect of low socioeconomic status during later life may induce pro-inflammatory responses and physiological changes that lead to dementia [54]. However, the mechanisms between gum treatment and cognitive health need further clinical trials and biochemical research to provide solid evidence.

Accounting for dental insurance coverage did more to reduce the association between dementia risk than accounting for the sociodemographic, behavioral, and health covariates. Having dental insurance is frequently considered a major driver for getting dental treatment, and the two are highly correlated [39]. Thus, it is not a surprise that controlling for dental insurance coverage would reduce the size of the association. Furthermore, we found that the protective effect of gum treatment on dementia is stronger in participants with higher income than those with lower income. One explanation is that socioeconomically advantaged older adults may have better access to dental care and oral health-related knowledge [39]. They may be able to accumulate socioeconomic resources, which can positively affect both oral health and cognitive function in later life. Additionally, the delayed divergence in dementia risk between the gum treatment and no gum treatment groups in the Kaplan-Meier plots may be explained by factors such as the gradual progression of periodontal disease, the time required for neuroprotective effects to manifest, and limitations of self-reported data on gum treatment. It is worth noting that our study is limited to a random subsample of HRS, and further studies on other large or national datasets are required to support the current findings.

Some limitations of this research need to be acknowledged. First, we assessed dementia on the basis of cognitive tests and proxy reports rather than clinical diagnoses. In prior studies using the HRS, researchers have demonstrated that the Langa-Weir algorithm correctly classified 74% and 86% of respondents, respectively [31]. Therefore, the concern of misclassification cannot be overlooked. Second, gum treatment was self-reported in the questionnaire, and the actual procedure was not differentiated in the study. Third, periodontal condition was not verified by oral examination. However, we adjusted for extensive sociodemographic, behavioral, and clinical factors to minimize confounding factors. Future studies should use dental treatment codes to determine gum treatment procedures and oral examinations to assess periodontal condition. Fourth, because the Dental Health Module was implemented in HRS 2008 and was not followed up, we used the gum treatment at baseline and did not account for the effect of gum treatment on periodontal health during the follow-up waves. Future research would benefit from more work with multiple waves of measures on dental procedures and periodontal health. Finally, it is important to recognize that the observed association between gum treatment and slower cognitive decline does not necessarily imply causation. Randomized controlled trials are needed to provide more definitive evidence for a causal relationship. However, our prospective cohort study design and rigorous adjustment for confounding factors strengthen the evidence for a potential causal link that warrants further investigation.

To the best of our knowledge, this is one of the first studies that found an association of gum treatment with slower cognitive decline and lower risk of dementia among older adults in the USA. These findings underscore the critical role of oral health in cognitive function, suggesting that gum treatment might be a feasible intervention for dementia prevention. It calls for greater integration of dental and medical care. Further research on maintaining good periodontal health and timely gum treatment may help establish effective dementia prevention strategies for older adults in the USA.

The Health and Retirement Study is sponsored by the National Institute on Aging (Grant No. U01AG009740) and is conducted by the University of Michigan. The authors thank all participants and staff for their contribution to this study.

The Health and Retirement Study (HRS) protocol was approved by the Institutional Review Board (IRB #HUM00061128) at the University of Michigan, and all participants provided written informed consent. This project is a secondary analysis of the HRS (<u>https://hrs.isr.umich.edu/about), which uses completely deidentified public datasets. As this study did not involve direct human participants and utilized only publicly available data, as per the Common Rule (45 CFR §46), Institutional Review Board review and approval were not required.

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

This study is partially supported by National Institutes of Health (NIH)/National Institute of Dental and Craniofacial Research (NIDCR) U01DE027512, NIH/National Institute of Aging (NIA) P30AG083257 and R56AG067619, and NIH/National Institute of Minority Health and Health Disparities (NIMHD) P50MD017356. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

Xiang Qi: conceptualization, methodology, formal analysis, investigation, software, and writing – original draft, review, and editing. Zheng Zhu, Katherine Wang, Yaguang Zheng, and An Li: methodology, formal analysis, and writing – original draft, review, and editing. Bei Wu: conceptualization, methodology, validation, investigation, data curation, writing – original draft, review, and editing, supervision, and funding acquisition.

The Health and Retirement Study (HRS) datasets are publicly available at the University of Michigan Institute for Social Research. Researchers may obtain the datasets after sending a data user agreement to the HRS team (https://hrs.isr.umich.edu/data-products).