Associations of Dental Health With the Progression of Hippocampal Atrophy in Community-Dwelling Individuals
The Ohasama Study
Abstract
Background and Objectives
Although tooth loss and periodontitis have been considered risk factors of Alzheimer disease, recent longitudinal researches have not found a significant association with hippocampal atrophy. Therefore, this study aimed to clarify a longitudinal association between the number of teeth present (NTP) and hippocampal atrophy dependent on the severity of periodontitis in a late middle-aged and older adult population.
Methods
This study included community-dwelling individuals aged 55 years or older who had no cognitive decline and had undergone brain MRI and oral and systemic data collection twice at 4-year intervals. Hippocampal volumes were obtained from MRIs by automated region-of-interest analysis. The mean periodontal probing depth (PD) was used as a measure of periodontitis. Multiple regression analysis was performed with the annual symmetric percentage change (SPC) of the hippocampal volume as the dependent variable and including an interaction term between NTP and mean PD as the independent variable. The interaction details were examined using the Johnson-Neyman technique and simple slope analysis. The 3-way interaction of NTP, mean PD, and time on hippocampal volume was analyzed using a linear mixed-effects model, and the interaction of NTP and time was examined in subgroups divided by the median mean PD. In all models, dropout bias was adjusted by inverse probability weighting.
Results
Data of 172 participants were analyzed. The qualitative interaction between NTP and the mean PD was significant for the annual SPC in the left hippocampus. The regression coefficient of the NTP on the annual SPC in the left hippocampus was positive (B = 0.038, p = 0.026) at the low-level mean PD (mean −1 SD) and negative (B = −0.054, p = 0.001) at the high-level mean PD (mean +1 SD). Similar results were obtained in the linear mixed-effects model; the interaction of NTP and time was significant in the higher mean PD group.
Discussion
In a late middle-aged and older cohort, fewer teeth were associated with a faster rate of left hippocampal atrophy in patients with mild periodontitis, whereas having more teeth was associated with a faster rate of atrophy in those with severe periodontitis. The importance of keeping teeth healthy is suggested.
The risk factors of dementia include diabetes mellitus and depression, but not all factors have been determined; only 40% of all cases with dementia are attributable to known modifiable risk factors.1 Tooth loss and periodontitis, caused by oral bacterial infection, have been recently suggested as risk factors of Alzheimer disease (AD) and related dementia, respectively.2,3 Tooth loss and periodontitis are highly prevalent worldwide,4 and if associated with the onset and progression of dementia, evaluation of their impact is important.
The clinical manifestations of AD are attributable to the progressive loss of neurons and synapses, with significant atrophy of memory-related structures in the medial temporal lobe (MTL) at an early stage, especially in the hippocampus5 and entorhinal cortex.6 Atrophy in the MTL, particularly in the hippocampus, has been identified as a potential biomarker for AD.7 Longitudinal studies have reported that many risk factors of AD are significantly associated with hippocampal atrophy.8-13 Regarding oral risk factors, animal studies have confirmed that a reduced number of teeth and the associated reduced masticatory activity cause hippocampal degeneration14 and that the chronic oral administration of periodontal bacteria induces neurodegeneration in the hippocampus of wild-type mice.15 Previous studies in humans have reported the relationship between (1) the number of teeth, whole-brain volume, and gray matter volume (GMV)16,17; (2) the number of teeth and left hippocampal volume in older adults with cognitive impairment18; and (3) people with edentulism and right hippocampus atrophy.19 However, these were cross-sectional studies. Although a longitudinal analysis recently reported that treatment of periodontitis improved AD-related brain atrophy,20 a previous study stated that the severity of periodontitis and tooth loss are not associated with morphological changes in the brain.21
In laboratory animals, tooth loss and periodontitis models are created as independent, and their interaction is not considered. However, in humans, tooth loss and periodontitis occur simultaneously in the oral cavity. In patients with mild periodontitis in each tooth, the increase in inflammation per increase in the number of teeth may be sufficiently small. Therefore, the decrease in masticatory function due to a smaller number of teeth is likely to cause brain atrophy.14 However, in those with severe periodontitis in each tooth, the increase in inflammation per tooth is non-negligible, and a large number of teeth may conversely cause brain atrophy owing to the increase in inflammation in the oral cavity.22 If this hypothesis is correct, then, the association between the number of teeth and changes in brain morphology depends on the degree of periodontitis. Therefore, a more detailed understanding of the relationship between the number of teeth and GMV changes in the human brain requires the use of statistical models that consider the interaction between the number of teeth and the degree of periodontitis.
This study aimed to evaluate the association between volume change rate in the hippocampus and the interaction of baseline number of teeth with the baseline periodontitis status (as a moderator variable) in a late middle-aged and older adult population living in the community, using longitudinal automated region-of-interest (ROI) and voxel-based morphometry (VBM) analysis.
Methods
Study Design
This study was conducted as part of the Ohasama Study, a prospective cohort study on hypertension and cardiovascular disease that began in 1986 in the general population of Ohasama, Iwate Prefecture, located in northern Japan.23,24 Further information on the Ohasama study is described in the eMethods (links.lww.com/WNL/C932). The inclusion criteria for these analyses were men and women aged 55 years or older who had at least 2 brain MRIs taken 4 years apart and had undergone an oral examination by dentists. The exclusion criteria were the presence of edentulism and suspected cognitive decline at baseline.
Participants
Figure 1 presents a flow diagram of the study participant selection procedure. We contacted all 3,147 Ohasama residents aged at least 55 years between 2009 and 2017. Overall, 1,156 residents wished to participate in the home blood pressure measurements, among whom 714 also wished to participate in additional MRI and dental examinations between April 2009 and 2017 (brain MRIs have been included in the database from April 2009). Among those residents, 296 had undergone at least 2 head MRI scans at 4-year intervals on record. We excluded 40 participants with missing dental data, 2 with missing medical data, and 34 with edentulism. Nineteen with baseline Mini-Mental State Examination (MMSE) scores ≤24 (maximum score: 30) points were excluded because of cognitive impairment. Eight participants with no MRI data (6 file open errors and 2 who did not provide consent for the use of MRI data) were excluded, and another 21 were excluded because preprocessing of their MRI findings was not successfully completed because of segmentation failure from bad tissue contrast. Finally, 172 participants (the follow-up group) were included in the analyses. The mean ± SD follow-up period was 4.0 ± 0.1 years. Conversely, of the 418 who did not undergo more than 1 MRI scan, 215 were considered as the non-follow-up group, after excluding 125 with missing data, 38 individuals with edentulism, and 40 with MMSE scores ≤24 points.
MMSE = Mini-Mental State Examination.
Data Collection
At the baseline, anthropometric measurements and blood samples were collected, and questionnaires on medical history, medication status, and smoking and drinking status were administered. Hypertension was defined as self-measured blood pressure (home blood pressure) of ≥135/85 mm Hg,25 antihypertensive drug use, or a history of hypertension. Diabetes was defined as a nonfasting blood glucose of ≥200 mg/dL, a glycated hemoglobin concentration of ≥6.5%, the use of diabetic medications, or a history of diabetes. Hypercholesterolemia was defined as a total cholesterol level of ≥220 mg/dL, antihypercholesterolemic drug use, or a history of hypercholesterolemia. History of cerebrovascular/cardiovascular disease was defined as brain stroke or ischemic heart disease and was confirmed by an interview, head MRI, or ECG. Drinking and smoking history was categorized as “never,” “former,” or “current.” The duration of education was classified as <10 years or ≥10 years. The body mass index (BMI) was calculated by dividing the weight in kilograms by the height in meters squared. Because the distribution of high-sensitivity C-reactive protein (hsCRP) measured by blood tests significantly deviated from the normal distribution, logarithmic transformation of hsCRP was used for the statistical analyses (log_hsCRP).26 Depressive symptoms were assessed using the Zung Self-Rating Depression Scale (SDS) in an interview.27 Cognitive functioning was assessed using the MMSE in an interview at the baseline and follow-up surveys.28
Oral Examination
Specially trained dentists counted the number of teeth present (NTP) and the periodontal probing depth (PD).29 The NTP was counted by excluding the remaining roots. Then, PD was measured at 4 points, 3 buccal and 1 palatal, on all teeth. The mean of all PD was used as the “mean PD.” The mean PD is an index of PD gain per tooth, and when multiplied by the NTP, it is considered to reflect the degree of periodontitis in the entire oral cavity. Because the mean PD itself was considered inappropriate as an indicator of the severity of periodontitis in the entire oral cavity, it was used only as a moderator variable for the NTP, and the main effect of the mean PD was excluded from the interpretation of the analysis results. Clinical attachment loss (CAL), often used in epidemiologic studies, reflects the cumulative history of past inflammation. Using CAL as the moderator of the NTP was inappropriate because only the cumulative history of inflammation around the present teeth was evaluated, ignoring the cumulative history of inflammation around the missing teeth before they were extracted.
MRI Acquisition and Preprocessing
All MRIs were captured using the EXCELART Vantage (1.5 T) (Toshiba Medical Systems, Tochigi, Japan) installed at the General Hanamaki Hospital (Iwate, Japan). The parameters of the axial FE3D sequence were as follows: repetition time, 14 milliseconds; echo time, 5.5 milliseconds; flip angle, 20°; 110 slices (slice thickness, 1.5 mm); matrix, 256 × 256; field of view, 220 mm; pixel size, 0.8594 × 0.8594 mm; and scan time, 4 minutes 59 seconds.
The MRIs were preprocessed using Statistical Parametric Mapping (SPM12, RRID: SCR_007037) and Computational Anatomy Toolbox for SPM (CAT12, RRID: SCR_019184) in MATLAB R2019b (Mathworks Corp., Natick, MA). Preprocessing details are described in eMethods (links.lww.com/WNL/C932).
ROI-Based Volume
To validate the volumetric changes in the left and right hippocampi, the “estimate mean values inside ROI” of CAT12 were used to calculate the GMV of the hippocampus of both baseline and follow-up scans. The percentage of volume change between the baseline and follow-up scans was calculated as the symmetric percentage change using the following equation: SPC: 200 × (follow-up GMV − baseline GMV)/(baseline GMV + follow-up GMV).30,31
Statistical Analyses
The Mann-Whitney U test and Fisher exact test were used to compare basic characteristics between the follow-up and no follow-up groups and between subgroups.
To adjust for any potential bias due to nonrandom dropout, we generated inverse probability weights (IPWs) that accounted for the probability of undergoing a second MRI scan.32 A binary indicator of the follow-up or no follow-up group was predicted by a logistic model with baseline variables. Weights were calculated as the inverse of the predicted probability of being in the follow-up group. Participants with characteristics with a higher probability of no follow-up had a greater weight in all the following statistical analyses.
In univariate analysis, a single regression analysis was performed using the annual SPC, which was calculated by dividing the SPC of the left and right hippocampi by the observation period (years) as the dependent variable. Age, sex, hypertension, diabetes, dyslipidemia, cerebral cardiovascular disease, smoking, alcohol consumption, education, BMI, log_hsCRP, SDS, baseline MMSE scores, NTP, and mean PD were each used as independent variables.
To analyze the association between hippocampal volume change and the interaction of NTP with mean PD, multiple regression analysis was performed using the annual SPC of the left and right hippocampi as the dependent variable. We fed both the NTP and mean PD as continuous variables into the statistical models: model 1, in which only age, sex, NTP, and mean PD were independent variables; model 2, in which the interaction term between NTP and mean PD was added to model 1; and model 3, in which all other variables were added to model 2. The interaction details were examined using the Johnson-Neyman technique33 and simple slope analysis.
Next, we performed 2 sensitivity analyses. First, to determine whether the interaction between the NTP and mean PD is also associated with changes in cognitive function, we conducted an analysis using the MMSE model, in which the dependent variable in model 3 was replaced by the annual change in the MMSE from baseline to follow-up ([follow-up MMSE − baseline MMSE]/observation period [years]). Adjusting for baseline cognitive status in regression analyses for cognitive change can introduce significant bias in the direction of the cross-sectional association between the baseline cognitive status and the baseline independent variable.34 Therefore, in the MMSE model, the baseline MMSE score was excluded from the independent variables for the analysis. Second, a linear mixed-effects model was used to analyze the 3-way interaction of time, NTP, and mean PD with the hippocampal volume as the dependent variable. Time was represented as years from baseline for each participant. This model was adjusted for the total intracranial volume (TIV) in addition to the same independent variables as in model 3 and each variable interacted with time. The continuous variables “age, time, BMI, MMSE, SDS, log_hsCRP, TIV, NTP, and mean PD” were mean centered. Because there were only 2 time points per participant, only random intercepts were included for each participant. We performed subgroup analysis to interpret the 3-way interaction, dividing the 2 groups by the median mean PD, and examined the interaction between the NTP and time.
Statistical analyses were performed using R4.2.1 (R Software for Statistical Computing, Vienna, Austria),35 and the “interactions” and “sjplot” packages were used to analyze and plot the interactions. The “lme4” and “LmerTest” packages were used for the linear mixed-effects model analysis. In all analyses, the significance level was set at 5%.
Longitudinal VBM
Longitudinal VBM was conducted within the hippocampal ROI to search for clusters with significant interaction between the NTP and mean PD associated with the 4-year GMV change rate. Methods for longitudinal VBM are described in the eMethods (links.lww.com/WNL/C932).
Standard Protocol Approvals, Registrations, and Patient Consents
This study was approved by the Institutional Review Board of the Teikyo University School of Medicine (approval number: 16-075-8) and the Tohoku University Tohoku Medical Megabank Organization (approval number: 2021-4-004) and conducted according to the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants.
Data Availability
Due to the sensitive nature of the data and ethical concerns, data from this cohort cannot be made available to other researchers.
Results
Study Sample
A comparison of the basic characteristics of the follow-up and no follow-up groups is listed in Table 1. Compared with the no follow-up group, the follow-up group had significantly higher SDS score and lower mean PD. In all subsequent statistical analyses, this bias was adjusted using IPW. The absolute standardized mean differences between the follow-up and no follow-up groups were <0.1, indicating a good covariate balance between the groups (eFigure 1, links.lww.com/WNL/C932).36
Hippocampal Volume
The ROI-based hippocampal volumes were as follows: left (2,564 [SD = 284] mm3) and right (2,897 [SD = 314] mm3) at baseline and left (2,462 [SD = 292] mm3) and right (2,783 [SD = 319] mm3) at follow-up (eFigure 2, links.lww.com/WNL/C932).
Univariate Analysis
Age and duration of education were significantly associated with the annual SPC in the left hippocampus, while age, cerebrovascular/cardiovascular disease, and duration of education were significantly associated with the annual SPC in the right hippocampus. The NTP and mean PD showed no association with the annual SPC in either hippocampus (eTable 1, links.lww.com/WNL/C932).
Multiple Regression Analysis With Continuous Variables
When multiple regression analysis was performed with the annual SPC of the left hippocampus as the dependent variable (Table 2), in model 1, only age was significant (R2 = 0.17, adjusted R2 = 0.15); in model 2, NTP, mean PD, and the interaction of NTP and mean PD became significant in addition to age (R2 = 0.26, adjusted R2 = 0.24). These variables were also significant in model 3 (R2 = 0.30, adjusted R2 = 0.23), in which various correction items were applied. When the annual SPC of the right hippocampus was used as the dependent variable (Table 3), age, mean PD, and the interaction of NTP and mean PD were significant in model 2, and only age was significant in models 1 and 3.
The plots of the interactions of the NTP and mean PD in model 3 with the annual SPCs of the left hippocampus as dependent variables are shown in Figure 2, A B. In the Johnson-Neyman plot that shows the size and significance of the regression coefficient throughout all observed levels of moderator variable, the regression coefficient of the NTP on the annual SPC in the left hippocampus had a significant positive coefficient when the mean PD was <2.17 and a significant negative coefficient when the mean PD was >3.18 (Figure 2A). When the mean PD was at 3 levels (mean, +1 SD and −1 SD), the regression coefficient of NTP on the annual SPC in the left hippocampus was significantly positive (B = 0.038, 95% CI 0.004–0.071, p = 0.026) at the low level (−1 SD) and significantly negative (B = −0.054, 95% CI −0.087 to −0.021, p = 0.001) at the high level (+1 SD) (Figure 2B).
(A) The Johnson-Neyman plot indicates the size and significance of the slope of NTP on annual symmetric percentage change of the left hippocampus throughout all observed levels of the mean PD. The shaded regions indicate 95% CIs. (B) Simple slope plot of the interaction between the NTP and mean PD on the annual symmetric percentage change of the left hippocampus for 4 levels (−1 SD, mean, +1 SD, and 4.5 mm) of the mean PD is shown. *Significant partial regression coefficient (p < 0.05). NTP = number of teeth present; PD = periodontal probing depth; SPC = symmetric percentage change.
Sensitivity Analysis
Because multiple regression analysis with continuous variables detected a significant interaction between the NTP and mean PD only in the left hippocampus, sensitivity analysis was performed for the left hippocampus only.
The results of the analysis of MMSE model are summarized in eTable 2 (links.lww.com/WNL/C932). The interaction between the NTP and mean PD was significant (p = 0.044), and the NTP (B = 0.042, 95% CI 0.007–0.077, p = 0.018) and mean PD were significantly associated with annual change in the MMSE score in addition to age, log_hsCRP, and SDS. In the Johnson-Neyman plot, the regression coefficient of the NTP on the annual change in MMSE had a significant positive coefficient when the mean PD was <2.69 mm (eFigure 3A). The regression coefficient of NTP on MMSE annual change significantly decreased as the mean PD increased, as did the regression coefficient of NTP on the annual SPC in the left hippocampus. When the mean PD was at 3 levels (mean, +1 SD and −1 SD), the regression coefficient of the NTP on the annual change in MMSE was significantly positive (B = 0.018, 95% CI 0.004–0.032, p = 0.012) only at the low level (−1 SD) (eFigure 3B).
The results of the linear mixed-effects model analysis are summarized in eTable 3 (links.lww.com/WNL/C932). Age, time, SDS, TIV, age × time, BMI × time, and 3-way interaction among NTP, mean PD, and time (B = −1.41, 95% CI −2.01 to −0.81, p < 0.001) were significantly associated with the left hippocampal volume. The basic characteristics of the 2 subgroups, divided by the median mean PD (2.60 mm), are summarized in Table 4. The lower mean PD group had a significantly higher NTP and a significantly higher percentage of drinkers and participants with more than 10 years of education compared with another group. The results of the subgroup analysis are summarized in eTables 4 and 5 and Figure 3. In the lower mean PD group, the interaction between the NTP and time was not significant (B = 0.61, 95% CI −0.34 to 1.56, p = 0.211); however, the regression coefficient of time on left hippocampal volume tended to decrease as the NTP decreased (Figure 3, A B). In the higher mean PD group, the interaction between the NTP and time was significant (B = −0.86, 95% CI −1.62 to −0.09, p = 0.031), with the regression coefficient of time on the left hippocampal volume significantly decreasing as the NTP increased (Figure 3, C D).
(A and C) The Johnson-Neyman plot indicates the size and significance of the slope of time on the left hippocampal volume throughout all observed levels of the NTP. The shaded regions indicate 95% CIs. (B and D) Simple slope plot of the interaction between time and the NTP on the left hippocampal volume for 3 levels (−1 SD, mean, +1 SD) of the NTP is shown. *Significant partial regression coefficient (p < 0.05). NTP = number of teeth present; PD = periodontal probing depth.
Longitudinal VBM
The results for the hippocampal ROI-based VBM are summarized in eTable 6 (links.lww.com/WNL/C932). Central portion of the left hippocampus was detected as a region where the negative interaction between the NTP and mean PD was significantly associated with the rate of GMV change (p-family-wise error = 0.035 cluster level). In a small region of the right hippocampus, a trend for a negative interaction between the NTP and mean PD, associated with the GMV change rate, was detected (p < 0.001 uncorrected voxel level, cluster size = 18 voxels).
Discussion
This study demonstrated that the interaction between the NTP and the severity of periodontitis is associated with morphological changes in the left hippocampus. In patients with mild periodontitis, the left hippocampus atrophied as the NTP decreased, while in patients with severe periodontitis, atrophy of the left hippocampus progressed as the NTP increased.
There was no previous analysis of the interaction between the NTP and mean PD, making it difficult to estimate its effect size. Therefore, no prior power analysis was performed, and all available samples were used as most observational studies do. The effect size of the interaction in this study (increase in R2 from model 1 to model 2: ΔR2 = 0.09) was large enough.
The follow-up group in this study had a significantly smaller mean PD value than that of the no follow-up group. Because the mean PD was one of the exposure factors in this study, it would have been appropriate to correct for dropout bias using IPW.
The hippocampal volumes calculated by manual tracing (left: 3,210 [SD = 397] mm3, right: 3,293 [SD = 424] mm337 and left: 2,165.71 [SE = 49.13] mm3, right: 2,141.06 [SE = 52.87] mm3),38 have been reported. The ROI-based hippocampal volume obtained in this study was reasonable compared with these previous reports.
In this study, the NTP had no association with the annual SPC of the left and right hippocampi in model 1, which did not include an interaction term. However, in model 2, which included an interaction term, the NTP (left hippocampus only), mean PD, and their interaction were statistically significant. In the left hippocampus only, this significance remained in model 3, which included many correction terms. This interaction was qualitative in that the direction of the association between the independent variable (NTP) and the dependent variable (annual SPC of the left hippocampus) was reversed depending on the value of the moderator variable (mean PD). These results support our hypothesis that when the degree of periodontitis of each tooth is mild, the association between the number of teeth and brain morphology is directly expressed; however, in cases of severe periodontitis of each tooth, the increase in inflammation per tooth may be not negligible. This interaction of the NTP and mean PD may explain why previous studies have been unable to show an association between periodontitis, the number of teeth, and brain morphological changes.21
Given the partial regression coefficient of age (B = −0.043) in model 3 for the left hippocampus (Table 2), the increase in the rate of atrophy of the left hippocampus due to 1 less tooth at the low-level mean PD (−1 SD: 2.05 mm) was equivalent to approximately 0.9 years of age (0.038/0.043 = 0.884); conversely, the increase in the rate of atrophy of the left hippocampus due to 1 more tooth at a high-level mean PD (+1 SD: 3.71 mm) was equivalent to approximately 1.3 years of age (0.054/0.043 = 1.256).
The results of the MMSE model suggested that cognitive decline was greater with fewer teeth in participants with mean PD < 2.69 mm. However, there was no significant association between the NTP and annual change in MMSE when the mean PD was ≥2.69 mm. Because having more teeth generally corresponds with younger age, cognitive decline due to a large number of teeth may be less detectable than hippocampal atrophy, which precedes cognitive decline.39
In the linear mixed-effects model, the 3-way interaction between the NTP, mean PD, and time was significantly associated with the left hippocampal volume. In the higher mean PD group, there was a significant negative interaction of the NTP and time on left hippocampal volume, confirming that the higher the number of teeth, the more progressive the left hippocampal atrophy over time. By contrast, in the lower mean PD group, the interaction between the NTP and time was not significant. However, there was a trend toward more progressive atrophy of the left hippocampus over time with a smaller number of teeth. The results of the linear mixed-effects model support the results of model 3.
The results of longitudinal VBM for the hippocampus support the influence of the interaction between the NTP and mean PD on the rate of volume change in the left hippocampus, even from a voxel-wise analysis, and suggest that a similar interaction may exist in a portion of the right hippocampus. Left hippocampal atrophy is purportedly greater than right hippocampal atrophy in patients with AD,40 and our VBM results support the possibility that the interaction between NTP and mean PD is associated with AD.
In this study, log_hsCRP showed no significant association with the annual SPC of the left hippocampus. In a previous study showing a cross-sectional association between chronic systemic inflammation and hippocampal volume, CRP ≥3 mg/L at ≥2 time points was defined as chronic low-grade inflammation.41 Periodontitis is a chronic inflammation and increases CRP in severe cases42; however, in this study, only 11 participants had CRP levels ≥3 mg/L. Porphyromonas gingivalis, a major pathogen of periodontitis, might cause temporary bacteremia even with daily activities, such as toothbrushing and chewing.43 Moreover, Porphyromonas gingivalis has been reported to invade the human brain and produce toxic proteases that increase amyloid-β production.44 Even periodontitis without elevated CRP might affect AD risk and hippocampal morphology.
The strength of this study is that it demonstrates that having more teeth may be a risk factor of dementia in patients with severe periodontitis. Considering that younger participants generally have more teeth even in the older adult group, our finding that the atrophy rate of the left hippocampus is greater with more teeth in the mean PD >3.18 mm is extremely important because it indicates that periodontitis may have a greater association with left hippocampal atrophy than the association exhibited by age.
As previous studies have suggested, it is important to preserve masticatory function and related neurologic activity by retaining teeth to maintain brain health. However, the results of this study suggest that retaining more teeth with severe periodontitis may promote hippocampal atrophy. Controlling periodontal disease progression through regular dental visits is crucial, whereas teeth with severe periodontitis may need to be extracted and replaced with appropriate prostheses.
This study has some limitations. First, the data used were extracted from a voluntary epidemiologic survey conducted in only 1 region of Japan. The dropout bias for the second MRI scan could be adjusted using IPW. However, the volunteer bias for the first MRI scan could not be corrected because of the lack of blood samples and dental data in those who did not participate in the first MRI scan. The small number of participants in this study (n = 172) may include a large proportion of health-conscious individuals and may not accurately reflect the characteristics of the general population of the region as a whole. As the comparison of the basic characteristics of the follow-up and no follow-up groups indicate, the prevalence of severe periodontitis particularly may be low and, therefore, larger studies are needed to increase generalizability. Second, we did not consider the changes in the number of teeth or periodontitis because the baseline survey was conducted. Third, all possible confounding factors could not be adjusted. Diet,45 lifestyle,46 statin,47 and ApoE4 genes48 have been suggested to be associated with cognitive and brain structure changes. However, the presence or absence of the ApoE4 gene was not investigated because the data used in this study were obtained from a cohort primarily targeting cardiovascular disease. We did not have data on diet, precise antihypercholesterolemic drug type, or lifestyle other than drinking and smoking history. Although diet and lifestyle are also associated with periodontitis,49,50 and the ApoE4 gene has been suggested to influence inflammation-induced hippocampal atrophy,41 we cannot rule out the possibility that significant confounding by these factors has not been corrected in this study. In addition, smoking history was analyzed using 3 category variables, and if a quantitative variable such as pack-years had been used, the results might be different. Future studies should verify our results using data from other cohorts, and analyses should be conducted that consider the effects of other factors.
In conclusion, this study revealed that having fewer teeth is associated with a faster rate of left hippocampal atrophy in patients with mild periodontitis, whereas having more teeth is associated with a faster rate of atrophy in those with severe periodontitis. This finding indicates that periodontitis may have a greater association with left hippocampal atrophy than the association exhibited by age. Furthermore, in cases of mild periodontitis, fewer teeth may be associated with a subsequent decline in cognitive function. These results highlight the importance of preserving the health of the teeth and not just retaining the teeth. Future studies should validate our findings using data from other cohorts.
Glossary
- AD
- Alzheimer disease
- BMI
- body mass index
- CAL
- clinical attachment loss
- GMV
- gray matter volume
- hsCRP
- high-sensitivity C-reactive protein
- IPW
- inverse probability weight
- log_hsCRP
- log-transformed hsCRP
- MMSE
- Mini-Mental State Examination
- MTL
- medial temporal lobe
- NTP
- number of teeth present
- PD
- periodontal probing depth
- ROI
- region of interest
- SDS
- Zung Self-Rating Depression Scale
- SPC
- symmetric percent change
- TIV
- total intracranial volume
- VBM
- voxel-based morphometry
Acknowledgment
The authors thank the residents and staff members in the Ohasama town and the staff members of the Hanamaki City Government, Iwate prefectural central hospital attachment Ohasama regional clinical Center (former Ohasama Hospital), General Hanamaki Hospital, for their valuable support on the Ohasama study project.
Appendix Authors
Footnote
CME Course: NPub.org/cmelist
Supplementary Material
File (supplementary_data1.pdf)
- Download
- 700.23 KB
References
1.
Livingston G, Huntley J, Sommerlad A, et al. Dementia prevention, intervention, and care: 2020 report of the Lancet Commission. Lancet. 2020;396(10248):413-446.
2.
Qi X, Zhu Z, Plassman BL, Wu B. Dose-response meta-analysis on tooth loss with the risk of cognitive impairment and dementia. J Am Med Dir Assoc. 2021;22(10):2039-2045.
3.
Kamer AR, Craig RG, Niederman R, Fortea J, de Leon MJ. Periodontal disease as a possible cause for Alzheimer's disease. Periodontol 2000. 2020;83(1):242-271.
4.
Peres MA, Macpherson LMD, Weyant RJ, et al. Oral diseases: a global public health challenge. Lancet. 2019;394(10194):249-260.
5.
Ferreira LK, Diniz BS, Forlenza OV, Busatto GF, Zanetti MV. Neurostructural predictors of Alzheimer's disease: a meta-analysis of VBM studies. Neurobiol Aging. 2011;32(10):1733-1741.
6.
Fennema-Notestine C, Hagler DJ Jr, McEvoy LK, et al. Structural MRI biomarkers for preclinical and mild Alzheimer's disease. Hum Brain Mapp. 2009;30(10):3238-3253.
7.
Chauveau L, Kuhn E, Palix C, et al. Medial temporal lobe subregional atrophy in aging and Alzheimer's disease: a longitudinal study. Front Aging Neurosci. 2021;13:750154.
8.
Alateeq K, Walsh EI, Cherbuin N. Higher blood pressure is associated with greater white matter lesions and brain atrophy: a systematic review with meta-analysis. J Clin Med. 2021;10(4):637.
9.
Xu W, Zhang C, Li JQ, et al. Age-related hearing loss accelerates cerebrospinal fluid tau levels and brain atrophy: a longitudinal study. Aging (Albany NY). 2019;11(10):3156-3169.
10.
Zou X, Durazzo TC, Meyerhoff DJ. Regional brain volume changes in alcohol-dependent individuals during short-term and long-term abstinence. Alcohol Clin Exp Res. 2018;42(6):1062-1072.
11.
Hawkins KA, Emadi N, Pearlson GD, et al. The effect of age and smoking on the hippocampus and memory in late middle age. Hippocampus. 2018;28(11):846-849.
12.
Arnone D, Job D, Selvaraj S, et al. Computational meta-analysis of statistical parametric maps in major depression. Hum Brain Mapp. 2016;37(4):1393-1404.
13.
Adachi Y, Ota K, Minami I, Yamada T, Watanabe T. Lower insulin secretion is associated with hippocampal and parahippocampal gyrus atrophy in elderly patients with type 2 diabetes mellitus. J Diabetes Investig. 2021;12(10):1908-1913.
14.
Weijenberg RA, Scherder EJ, Lobbezoo F. Mastication for the mind: the relationship between mastication and cognition in ageing and dementia. Neurosci Biobehav Rev. 2011;35(3):483-497.
15.
Ilievski V, Zuchowska PK, Green SJ, et al. Chronic oral application of a periodontal pathogen results in brain inflammation, neurodegeneration and amyloid beta production in wild type mice. PLoS One. 2018;13(10):e0204941.
16.
Dintica CS, Rizzuto D, Marseglia A, et al. Tooth loss is associated with accelerated cognitive decline and volumetric brain differences: a population-based study. Neurobiol Aging. 2018;67:23-30.
17.
Egashira R, Umezaki Y, Mizutani S, et al. Relationship between cerebral atrophy and number of present teeth in elderly individuals with cognitive decline. Exp Gerontol. 2021;144:111189.
18.
Lin CS, Lin HH, Fann SW, et al. Association between tooth loss and gray matter volume in cognitive impairment. Brain Imaging Behav. 2020;14(2):396-407.
19.
Kobayashi T, Kubota M, Takahashi T, et al. Effects of tooth loss on brain structure: a voxel-based morphometry study. J Prosthodont Res. 2018;62(3):337-341.
20.
Schwahn C, Frenzel S, Holtfreter B, et al. Effect of periodontal treatment on preclinical Alzheimer's disease-results of a trial emulation approach. Alzheimers Dement. 2022;18(1):127-141.
21.
Adam HS, Lakshminarayan K, Wang W, et al. The prospective association between periodontal disease and brain imaging outcomes: the atherosclerosis risk in communities study. J Clin Periodontol. 2022;49(4):322-334.
22.
Gu Y, Vorburger R, Scarmeas N, et al. Circulating inflammatory biomarkers in relation to brain structural measurements in a non-demented elderly population. Brain Behav Immun. 2017;65:150-160.
23.
Imai Y, Nagai K, Sakuma M, et al. Ambulatory blood pressure of adults in Ohasama, Japan. Hypertension. 1993;22(6):900-912.
24.
Ohkubo T, Asayama K, Kikuya M, et al. How many times should blood pressure be measured at home for better prediction of stroke risk? Ten-year follow-up results from the Ohasama study. J Hypertens. 2004;22(6):1099-1104.
25.
Tsuji I, Imai Y, Nagai K, et al. Proposal of reference values for home blood pressure measurement: prognostic criteria based on a prospective observation of the general population in Ohasama, Japan. Am J Hypertens. 1997;10(4 pt 1):409-418.
26.
Taki Y, Thyreau B, Kinomura S, et al. Correlation between high-sensitivity c-reactive protein and brain gray matter volume in healthy elderly subjects. Hum Brain Mapp. 2013;34(10):2418-2424.
27.
Zung WW. Depression in the normal aged. Psychosomatics. 1967;8(5):287-292.
28.
Folstein MF, Folstein SE, McHugh PR. “Mini-mental state.” A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res. 1975;12(3):189-198.
29.
Yamada S, Komiyama T, Ohi T, et al. Regular dental visits, periodontitis, tooth loss, and atherosclerosis: the Ohasama study. J Periodontal Res. 2022;57(3):615-622.
30.
Rosas HD, Reuter M, Doros G, et al. A tale of two factors: what determines the rate of progression in Huntington's disease? A longitudinal MRI study. Mov Disord. 2011;26(9):1691-1697.
31.
Berry DA, Ayers GD. Symmetrized percent change for treatment comparisons. Am Stat. 2006;60(1):27-31.
32.
Rinsky JL, Richardson DB, Wing S, et al. Assessing the potential for bias from nonresponse to a study follow-up interview: an example from the agricultural health study. Am J Epidemiol. 2017;186(4):395-404.
33.
Johnson PO, Fay LC. The Johnson-Neyman technique, its theory and application. Psychometrika. 1950;15(4):349-367.
34.
Glymour MM, Weuve J, Berkman LF, Kawachi I, Robins JM. When is baseline adjustment useful in analyses of change? An example with education and cognitive change. Am J Epidemiol. 2005;162(3):267-278.
35.
R: A Language and Environment for Statistical Computing [Computer Program]. R Foundation for Statistical Computing; 2021.
36.
Stuart EA, Lee BK, Leacy FP. Prognostic score-based balance measures can be a useful diagnostic for propensity score methods in comparative effectiveness research. J Clin Epidemiol. 2013;66(8 suppl):S84-S90.
37.
Schmidt MF, Storrs JM, Freeman KB, et al. A comparison of manual tracing and Freesurfer for estimating hippocampal volume over the adult lifespan. Hum Brain Mapp. 2018;39(6):2500-2513.
38.
Zierhut KC, Grassmann R, Kaufmann J, Steiner J, Bogerts B, Schiltz K. Hippocampal CA1 deformity is related to symptom severity and antipsychotic dosage in schizophrenia. Brain. 2013;136(pt 3):804-814.
39.
Martin SB, Smith CD, Collins HR, Schmitt FA, Gold BT. Evidence that volume of anterior medial temporal lobe is reduced in seniors destined for mild cognitive impairment. Neurobiol Aging. 2010;31(7):1099-1106.
40.
Minkova L, Habich A, Peter J, Kaller CP, Eickhoff SB, Kloppel S. Gray matter asymmetries in aging and neurodegeneration: a review and meta-analysis. Hum Brain Mapp. 2017;38(12):5890-5904.
41.
Tao Q, Ang TFA, DeCarli C, et al. Association of chronic low-grade inflammation with risk of Alzheimer disease in APOE4 carriers. JAMA Netw Open. 2018;1(6):e183597.
42.
Machado V, Botelho J, Escalda C, et al. Serum c-reactive protein and periodontitis: a systematic review and meta-analysis. Front Immunol. 2021;12:706432.
43.
Forner L, Larsen T, Kilian M, Holmstrup P. Incidence of bacteremia after chewing, tooth brushing and scaling in individuals with periodontal inflammation. J Clin Periodontol. 2006;33(6):401-407.
44.
Dominy SS, Lynch C, Ermini F, et al. Porphyromonas gingivalis in Alzheimer's disease brains: evidence for disease causation and treatment with small-molecule inhibitors. Sci Adv. 2019;5(1):eaau3333.
45.
Otsuka R, Nishita Y, Nakamura A, et al. Dietary diversity is associated with longitudinal changes in hippocampal volume among Japanese community dwellers. Eur J Clin Nutr. 2021;75(6):946-953.
46.
Kivipelto M, Mangialasche F, Ngandu T. Lifestyle interventions to prevent cognitive impairment, dementia and Alzheimer disease. Nat Rev Neurol. 2018;14(11):653-666.
47.
Poly TN, Islam MM, Walther BA, et al. Association between use of statin and risk of dementia: a meta-analysis of observational studies. Neuroepidemiology. 2020;54(3):214-226.
48.
Serrano-Pozo A, Das S, Hyman BT. APOE and Alzheimer's disease: advances in genetics, pathophysiology, and therapeutic approaches. Lancet Neurol. 2021;20(1):68-80.
49.
Cassiano LS, Peres MA, Motta JVS, et al. Periodontitis is associated with consumption of processed and ultra-processed foods: findings from a population-based study. Nutrients. 2022;14(18):3735.
50.
Chapple IL, Bouchard P, Cagetti MG, et al. Interaction of lifestyle, behaviour or systemic diseases with dental caries and periodontal diseases: consensus report of group 2 of the joint EFP/ORCA workshop on the boundaries between caries and periodontal diseases. J Clin Periodontol. 2017;44(suppl 18):S39-S51.
Information & Authors
Information
Published In
Copyright
Copyright © 2023 The Author(s). Published by Wolters Kluwer Health, Inc. on behalf of the American Academy of Neurology. This is an open access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND), which permits downloading and sharing the work provided it is properly cited. The work cannot be changed in any way or used commercially without permission from the journal.
Publication History
Received: January 25, 2023
Accepted: May 10, 2023
Published online: July 5, 2023
Published in print: September 5, 2023
Disclosure
K. Asayama, H. Metoki, Y. Imai, and T. Ohkubo concurrently held the position of director of the Tohoku Institute for Management of Blood Pressure, supported by Omron Healthcare Co., Ltd. K. Asayama and T. Ohkubo received a joint research grant from Omron Healthcare Co., Ltd. S. Yamaguchi, T. Murakami, M. Satoh, T. Komiyama, T. Ohi, Y. Miyoshi, K. Endo, T. Hiratsuka, A. Hara, Y. Tatsumi, T. Totsune, M. Kikuya, K. Nomura, A. Hozawa, M. Watanabe, and Y. Hattori have nothing to disclose. Go to Neurology.org/N for full disclosures.
Study Funding
This study was supported by Grants for Scientific Research, Ministry of Education, Culture, Sports, Science and Technology, Japan (grant numbers 18K09674, 18K09904, 18K17396, 19K19466, 19H03908, 19K10662, 20K08612, 21K09970, 21K10452, 21K10478, 21H04854, 21K17313, 21K19670, and 22H03358); the internal research grants from Keio University; the Japan Arteriosclerosis Prevention Fund; grant-in-aid from the Ministry of Health, Labor, and Welfare, Japan (grant numbers H29–Junkankitou–Ippan–003 and 20FA1002); ACRO Incubation Grants of Teikyo University; the Academic Contributions from Pfizer Japan Inc. and Bayer Yakuhin, Ltd.; scholarship donations from Chugai Pharmaceutical Co., Ltd., Daiichi Sankyo Co., Ltd.; research support from Astellas Pharma Inc. and Takeda Pharmaceutical Co., Ltd.; The Health Care Science Institute Research Grant; Health Science Center Research Grant; and Takeda Science Foundation.
Authors
Metrics & Citations
Metrics
Citation information is sourced from Crossref Cited-by service.
Citations
Download Citations
If you have the appropriate software installed, you can download article citation data to the citation manager of your choice. Select your manager software from the list below and click Download.
Cited By
- No genetic causal associations between periodontitis and brain atrophy or cognitive impairment: evidence from a comprehensive bidirectional Mendelian randomization study, BMC Oral Health, 24, 1, (2024).https://doi.org/10.1186/s12903-024-04367-7
- Mediating and moderating effects of plasma proteomic biomarkers on the association between poor oral health problems and brain white matter microstructural integrity: the UK Biobank study, Molecular Psychiatry, (2024).https://doi.org/10.1038/s41380-024-02678-3
- Restoring periodontal tissue homoeostasis prevents cognitive decline by reducing the number of Serpina3nhigh astrocytes in the hippocampus, The Innovation, 5, 1, (100547), (2024).https://doi.org/10.1016/j.xinn.2023.100547
- Research Trends in the Comorbidity Between Periodontitis and Neurodegenerative Diseases, International Dental Journal, (2024).https://doi.org/10.1016/j.identj.2024.07.1212
- On some challenges of psychological research in late adulthood and aging, Current Opinion in Psychology, 55, (101745), (2024).https://doi.org/10.1016/j.copsyc.2023.101745
- Periodontitis and brain magnetic resonance imaging markers of Alzheimer's disease and cognitive aging, Alzheimer's & Dementia, 20, 3, (2191-2208), (2024).https://doi.org/10.1002/alz.13683
- Microglial DBP Signaling Mediates Behavioral Abnormality Induced by Chronic Periodontitis in Mice, Advanced Science, 11, 46, (2024).https://doi.org/10.1002/advs.202406269
- Advances in the prevention and treatment of Alzheimer’s disease based on oral bacteria, Frontiers in Psychiatry, 14, (2023).https://doi.org/10.3389/fpsyt.2023.1291455
Loading...
View Options
View options
PDF and All Supplements
Download PDF and Supplementary MaterialShort Form
View Short FormFull Text
View Full TextLogin options
Check if you have access through your login credentials or your institution to get full access on this article.
Personal login Institutional LoginPurchase Options
The neurology.org payment platform is currently offline. Our technical team is working as quickly as possible to restore service.
If you need immediate support or to place an order, please call or email customer service:
- 1-800-638-3030 for U.S. customers - 8:30 - 7 pm ET (M-F)
- 1-301-223-2300 for customers outside the U.S. - 8:30 - 7 pm ET (M-F)
- [email protected]
We appreciate your patience during this time and apologize for any inconvenience.