ABSTRACT
Aims
To evaluate the longitudinal association between periodontitis and stroke onset using medical claims data and also to assess the relationship between periodontitis and stroke‐related medical costs and length of hospital stay.
Materials and Methods
In total, 1997 individuals aged 75 or 80 years were included. The association between periodontitis and stroke onset over a 5‐year period was evaluated using a Cox proportional hazards model, with stroke onset as the dependent variable and periodontitis as the independent variable. Restricted mean survival time (RMST) was compared across periodontitis severity groups. In addition, generalised linear models were used to evaluate the relationships of periodontitis with stroke‐related medical costs and length of hospital stay among participants who experienced a stroke during the follow‐up period.
Results
Participants with severe periodontitis had a significantly higher hazard ratio for stroke onset after 23 months of follow‐up compared to those without periodontitis and exhibited a significantly reduced RMST for stroke onset. Among 80‐year‐old participants, those with periodontitis incurred significantly higher treatment costs and spent more days in health care compared to participants without periodontitis. No significant associations were observed among 75‐year‐old participants.
Conclusion
Our findings suggest that periodontitis may be associated with the incidence and severity of stroke, potentially impacting the increased medical care costs associated with stroke.
Keywords: cohort study, healthcare cost, hospitalisation, periodontitis, stroke
1. Introduction
Stroke is a major manifestation of atherosclerotic disease and a collective term encompassing intracerebral haemorrhage, subarachnoid haemorrhage and cerebral infarction. In Japan, mortality from stroke has declined and accounts for a smaller proportion of all‐cause deaths than cancer or heart disease (Ministry of Health, Labour and Welfare 2024). In contrast, stroke‐related mortality is increasing globally; in 2021, approximately 7.3 million people died from stroke (GBD 2021 Stroke Risk Factor Collaborators 2024). Moreover, stroke frequently results in persistent disabilities, including motor impairment, sensory impairment and higher cortical dysfunction, making it a leading cause of long‐term care dependency (Ministry of Health, Labour and Welfare 2022). The incidence of stroke increases with age and peaks in individuals in their 70s (Toyoda et al. 2019). In Japan, where the population is rapidly ageing, stroke prevention is critically important for maintaining the quality of life in older adults and reducing social security expenditures, including medical and long‐term care costs.
Previous studies have suggested that periodontal disease may increase the risk of stroke (Dietrich et al. 2013; El Masri et al. 2024; Lafon et al. 2014). Periodontal pathogens and their toxins can enter the bloodstream from periodontal tissues and may induce inflammation in blood vessels distant from the oral cavity (Schenkein et al. 2020). In addition, chronic inflammation associated with periodontal disease has been reported to affect other tissues, and through these mechanisms, periodontal disease may increase the risk of cardiovascular disease (Schenkein et al. 2020; Villoria et al. 2024). However, although numerous studies have investigated the association between periodontal disease and cardiovascular disease, studies focusing on its association with stroke remain limited and are insufficient to draw definitive conclusions (Fagundes et al. 2019; Leira et al. 2017). A recent review based on retrospective cohort studies concluded that periodontal disease increases the risk of stroke onset (El Masri et al. 2024), whereas a Mendelian randomisation study suggested that the observed association may be attributable to confounding factors (Bell et al. 2020). Furthermore, several review articles have not identified a causal relationship between periodontal disease and stroke (Asmat‐Abanto et al. 2025; Meng and Chen 2025).
Stroke often requires hospitalisation and surgical intervention, resulting in substantial medical costs. Although several studies have reported an association between periodontal disease and the incidence of stroke, no study has evaluated the relationship between periodontal disease and stroke‐related medical costs or length of hospital stay. Clarifying the extent to which periodontal disease influences stroke‐related medical costs and length of hospital stay will emphasise the importance of maintaining oral health, in order to reduce medical expenditures and hospitalisation burden through stroke prevention.
We evaluated the longitudinal association between periodontal status and stroke onset using medical claims data from Japanese late‐stage older adults. In addition, we assessed the association between periodontal status and stroke‐related medical costs and length of hospital stay among individuals who experienced a stroke during the follow‐up period.
2. Method
2.1. Participants
We used data from dental check‐ups conducted by a public health insurance provider as part of health programmes in 2014, together with medical claims data from fiscal years 2014–2019 (Saito et al. 2021). The participants in this dental check‐up were residents of Mie prefecture who were either 75 or 80 years old. As periodontal status cannot be accurately assessed in individuals with a limited number of teeth, participants with ≥ 20 teeth were included. Participants who reported a medical history of stroke or heart disease on the baseline questionnaire and those whose medical claims data confirmed visits to medical institutions for stroke during fiscal year 2014 were excluded. The dental check‐ups were conducted as part of routine periodic health examinations for the older population rather than for research purposes; therefore, individual informed consent for use of the data was not obtained. Permission to use de‐identified data was obtained from the insurer. The Institutional Review Board of Aichi Gakuin University, School of Dentistry, approved the study protocol (approval no. 777). The study was performed in accordance with the World Medical Association Declaration of Helsinki. The study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for observational studies in humans.
2.2. Dental Check‐Up
Individuals who elected to undergo a dental check‐up visited dental clinics participating in the Mie Dental Association. Tooth and periodontal status were assessed during the dental check‐ups. The Community Periodontal Index (CPI) was used to assess periodontal status, as this method is commonly used in public oral health examinations in Japan (World Health Organisation 1997).
2.3. Questionnaire
Data on height, weight, smoking status (yes/no) and medical history (stroke, diabetes mellitus and heart disease) were obtained using a self‐administered questionnaire.
2.4. Medical Claims Data
Data on the month of stroke onset, presence of hypertension, total medical costs in the baseline year, stroke‐related medical costs, length of hospital stay during the follow‐up period and economic status were obtained from medical claims data. Stroke onset was identified by longitudinally tracking participants using medical claims data from the baseline fiscal year 2014 through the end of fiscal year 2019. For details on the method, refer to Supporting Information: Appendix A.
2.5. Statistical Analysis
Periodontitis was classified into three categories based on the individual CPI code: codes 0, 1 and 2 were defined as no periodontitis, code 3 as moderate periodontitis and code 4 as severe periodontitis. Economic status was categorised according to resident tax status: participants exempt from taxation were classified as having low income, whereas those subject to taxation were classified as having normal income. Body mass index (BMI; kg/m2) was calculated using height (m) and weight (kg). For stroke‐related medical costs, cumulative costs tend to be higher among participants with earlier stroke onset; therefore, monthly medical costs were calculated by dividing cumulative costs by the post‐onset follow‐up period and used in the analyses.
Regarding the associations between stroke onset and individual variables, the χ 2 test was used to assess differences in proportions, and the t‐test was used to evaluate differences in means. The survival curves for stroke onset crossed among the three periodontitis groups, indicating that the proportional hazards assumption was not satisfied over the entire follow‐up period (Figure 1). Therefore, two analytical approaches were used to evaluate the association between periodontitis and stroke onset. First, the observation period was divided at the 23rd month, when the survival curves crossed, and analyses were conducted separately for two intervals, from the start of follow‐up to the 22nd month and from the 23rd month to the 60th month, using Cox proportional hazards models. Cox proportional hazards models were constructed with stroke onset as the dependent variable and age, sex, medical history (diabetes mellitus and hypertension), smoking status, economic status, periodontitis status, BMI and total medical costs as independent variables. Hazard ratios (HRs) and 95% confidence intervals (CIs) were calculated. Participants who died or lost insurance coverage during the follow‐up period were censored at the time of discontinuation. To verify the robustness of the results, we conducted sensitivity analyses using multiple landmark time points. We set the landmark times at 1, 2, 3 and 4 years after the start of observation and analysed the association between periodontitis and stroke during the periods before and after these time points using Cox proportional hazards models. Second, restricted mean survival time (RMST) was used to evaluate the association between periodontitis and stroke onset. RMST represents the mean survival time over the observation period, and comparisons of RMST allow summarisation of the effects of exposure variables. This approach permits the analysis of survival data without assuming proportional hazards. RMST was estimated using survival curves derived from the Cox proportional hazards models described above. Differences in RMST, the ratio of RMST, the ratio of restricted mean lost time (RMLT) and corresponding 95% CIs were calculated across the three periodontitis groups.
FIGURE 1.
Survival curves for stroke onset among the three periodontitis groups.
Moreover, the associations between periodontitis and stroke‐related medical costs and length of hospital stay were evaluated among participants who experienced a stroke during the follow‐up period. Generalised linear models (distribution: Tweedie; link function: log; robust standard errors) were used, with stroke‐related medical costs per month or stroke‐related length of hospital stay as the dependent variable and sex, medical history, smoking status, economic status, periodontitis, BMI and total medical costs as independent variables. Analyses of stroke‐related medical costs and length of hospital stay were conducted separately. In the analysis of periodontitis and stroke‐related medical costs, the post‐stroke follow‐up period was included as an additional independent variable. Statistical analyses were performed using SPSS (version 28.0; IBM Corp., Armonk, NY, USA), and RMST analyses were conducted using R (version 4.2.1; R Foundation for Statistical Computing, Vienna, Austria). p‐values < 0.05 were considered statistically significant.
3. Results
Of the 33,378 individuals eligible for dental check‐ups in fiscal year 2014, 4984 underwent dental exams (attendance rate, 14.9%). The analysis included 1997 participants after excluding 26 individuals who were no longer insured by 31 March 2015; 355 who had visited medical institutions for stroke in fiscal year 2014; 370 with a history of stroke; 925 with a history of heart disease; 2052 with < 20 teeth; and 55 with missing data (Figure 2). During the follow‐up period, 232 individuals (11.6%) experienced stroke.
FIGURE 2.
Participant summary and flowchart for the follow‐up period.
Table 1 presents the association between stroke onset and each variable. Participants with severe periodontitis had a higher proportion of stroke onset; however, this difference was not statistically significant. No other variables were significantly associated with stroke onset.
TABLE 1.
Relationship between the incidence of stroke and independent variables.
| Independent variable | Incidence of stroke | p | |
|---|---|---|---|
| No | Yes | ||
| n (%) | |||
| Total | 1765 (88.4) | 232 (11.6) | |
| Age | |||
| 75 years | 1171 (88.0) | 160 (12.0) | 0.459 |
| 80 years | 594 (89.2) | 72 (10.8) | |
| Sex | |||
| Men | 749 (87.2) | 110 (12.8) | 0.159 |
| Women | 1106 (89.3) | 122 (10.7) | |
| Medical history | |||
| Diabetes mellitus | |||
| No | 1584 (88.3) | 209 (11.7) | 1.000 |
| Yes | 181 (88.7) | 23 (11.3) | |
| Hypertension | |||
| No | 1140 (88.4) | 149 (11.6) | 0.942 |
| Yes | 625 (88.3) | 83 (11.7) | |
| Smoking status | |||
| No | 1712 (88.4) | 225 (11.6) | 1.000 |
| Yes | 53 (88.3) | 7 (11.7) | |
| Economic status | |||
| Normal | 1314 (88.6) | 169 (11.4) | 0.632 |
| Low | 451 (87.7) | 63 (12.3) | |
| Periodontitis | |||
| No | 697 (89.4) | 83 (10.6) | 0.108 |
| Moderate | 716 (88.9) | 89 (11.1) | |
| Severe | 352 (85.4) | 60 (14.6) | |
| Mean (SD) | |||
| Body mass index (kg/m2) | 22.3 (3.0) | 22.4 (2.7) | 0.781 |
| Medical costs at baseline year (×10,000 yen) | 33.3 (59.7) | 38.6 (56.0) | 0.201 |
Abbreviation: SD, standard deviation.
Table 2 summarises the results of the Cox proportional hazards models, with stroke onset as the outcome. During the first 22 months of follow‐up, economic status and BMI were significantly associated with stroke onset, whereas periodontitis was not. From month 23 onward, periodontitis and total medical costs were significantly associated with stroke onset. Participants with severe periodontitis had a significantly higher HR for stroke onset compared to those without periodontitis (HR, 1.65; 95% CI: 1.08–2.53). In sensitivity analyses using three landmark times, the HRs for stroke were significantly higher among participants with severe periodontitis than in those without periodontitis at 1, 2 and 3 years or later (Tables S1–S3). For the period starting from the fourth year onward, participants with severe periodontitis had higher HR for stroke, showing a trend similar to that observed in analyses at other landmark time points; however, this was not statistically significant (Table S4).
TABLE 2.
Relationship between the incidence of stroke and independent variables during follow‐up period by Cox proportional hazards models.
| Independent variable | Incidence of stroke (0–22 months) | Incidence of stroke (23–60 months) | ||||||
|---|---|---|---|---|---|---|---|---|
| Crude HR (95% CI) | p | Adjusted HR (95% CI) | p | Crude HR (95% CI) | p | Adjusted HR (95% CI) | p | |
| Age | ||||||||
| 75 years | 1 | 1 | 1 | 1 | ||||
| 80 years | 0.89 (0.57–1.39) | 0.605 | 1.07 (0.67–1.71) | 0.784 | 1.04 (0.73–1.49) | 0.838 | 0.99 (0.69–1.43) | 0.971 |
| Sex | ||||||||
| Men | 1 | 1 | 1 | 1 | ||||
| Women | 1.08 (0.73–1.61) | 0.692 | 1.02 (0.66–1.58) | 0.925 | 0.77 (0.55–1.08) | 0.127 | 0.78 (0.55–1.12) | 0.181 |
| Medical history | ||||||||
| Diabetes mellitus | ||||||||
| No | 1 | 1 | 1 | 1 | ||||
| Yes | 1.11 (0.48–2.55) | 0.807 | 1.34 (0.55–3.23) | 0.518 | 1.29 (0.77–2.14) | 0.333 | 1.19 (0.71–1.99) | 0.520 |
| Hypertension | ||||||||
| No | 1 | 1 | 1 | 1 | ||||
| Yes | 0.87 (0.57–1.32) | 0.503 | 0.83 (0.53–1.30) | 0.412 | 1.03 (0.73–1.47) | 0.854 | 1.05 (0.74–1.50) | 0.776 |
| Smoking status | ||||||||
| No | 1 | 1 | 1 | 1 | ||||
| Yes | 1.42 (0.45–4.48) | 0.552 | 1.44 (0.45–4.68) | 0.541 | 0.96 (0.35–2.59) | 0.929 | 1.07 (0.39–2.94) | 0.896 |
| Economic status | ||||||||
| Normal | 1 | 1 | 1 | 1 | ||||
| Low | 1.41 (0.92–2.18) | 0.119 | 1.65 (1.03–2.64) | 0.039 | 0.96 (0.65–1.42) | 0.818 | 1.06 (0.70–1.59) | 0.798 |
| Periodontitis | ||||||||
| No | 1 | 1 | 1 | 1 | ||||
| Moderate | 1.15 (0.73–1.80) | 0.547 | 1.07 (0.67–1.72) | 0.781 | 1.01 (0.67–1.51) | 0.982 | 1.02 (0.68–1.53) | 0.940 |
| Severe | 0.89 (0.52–1.53) | 0.672 | 0.85 (0.48–1.51) | 0.581 | 1.67 (1.10–2.54) | 0.018 | 1.65 (1.08–2.53) | 0.021 |
| Body mass index (kg/m2) | 1.10 (1.03–1.18) | 0.007 | 1.13 (1.05–1.22) | < 0.001 | 1.02 (0.96–1.08) | 0.605 | 1.00 (0.95–1.06) | 0.944 |
| Medical costs at baseline year (×10,000 yen) | 1.00 (0.99–1.00) | 0.053 | 1.00 (0.99–1.00) | 0.057 | 1.00 (1.00–1.01) | 0.012 | 1.00 (1.00–1.01) | 0.013 |
Abbreviations: CI, confidence interval; HR, hazard ratio.
Table 3 presents the differences in RMST, the ratios of RMST and the ratios of RMLT among the three periodontitis groups. Compared to participants without periodontitis, those with severe periodontitis had a significantly shorter RMST for stroke onset. The difference in RMST was ˗0.80 (95% CI: −1.21 to –0.39), the ratio of RMST was 0.99 (95% CI: 0.98–0.99) and the ratio of RMLT was 1.22 (95% CI: 1.09–1.37). Compared to participants with moderate periodontitis, those with severe periodontitis had a significantly shorter RMST for stroke onset. The difference in RMST was –0.61 (95% CI: –1.03 to –0.18), the ratio of RMST was 0.99 (95% CI: 0.98–1.00) and the ratio of RMLT was 1.17 (95% CI: 1.04–1.30). No significant difference in RMST was observed between participants without periodontitis and those with moderate periodontitis.
TABLE 3.
Analysis of RMST of the incidence of stroke due to periodontitis.
| Difference of RMST a (95% CI) | p | Raito of RMST a (95% CI) | p | Ratio of RMTL a (95% CI) | p | |
|---|---|---|---|---|---|---|
| Periodontitis | ||||||
| No vs. severe | −0.80 (−1.21, −0.39) | < 0.001 | 0.99 (0.98, 0.99) | < 0.001 | 1.22 (1.09, 1.37) | < 0.001 |
| No vs. moderate | −0.24 (−0.730, 0.24) | 0.321 | 1.00 (0.99, 1.00) | 0.321 | 1.07 (0.94, 1.22) | 0.321 |
| Moderate vs. severe | −0.61 (−1.03, −0.18) | 0.005 | 0.99 (0.98, 1.00) | 0.005 | 1.17 (1.04, 1.30) | 0.006 |
Abbreviations: CI, confidence interval; RMST, restricted mean survival time; RMTL, restricted mean time lost.
Adjusted for age, sex, medical history of diabetes and hypertension, smoking status, economic status, body mass index and medical costs at baseline year.
Table 4 presents the results of age‐specific generalised linear models among participants who developed a stroke, with stroke‐related medical costs or length of hospital stay as the dependent variable. Among participants who developed a stroke, 160 were aged 75 years and 72 were aged 80 years. Among 80‐year‐old participants, those with moderate or severe periodontitis had significantly higher cost ratios for stroke‐related medical costs compared to those without periodontitis. The cost ratios were 4.18 (95% CI: 1.71–10.20) and 6.32 (95% CI: 2.21–18.06), respectively. For the analysis of length of hospital stay among 80‐year‐old participants, only 1 participant was a smoker, and the model did not converge; therefore, the analysis was conducted in the remaining 71 participants after excluding the smoker. Among 80‐year‐old participants, those with moderate or severe periodontitis had significantly higher day ratios for stroke‐related length of hospital stay compared to those without periodontitis; the day ratios were 7.14 (95% CI: 2.18–23.42) and 21.85 (95% CI: 5.94–80.35), respectively. No significant associations between periodontitis and stroke‐related medical costs or length of hospital stay were observed among 75‐year‐old participants.
TABLE 4.
Relationship of stroke‐related medical costs and length of hospital stay with periodontitis according to age by generalised linear models.
| Independent variables a | Dependent variable: Stroke related medical costs per month | Dependent variable: Stroke related length of hospital stay | ||||||
|---|---|---|---|---|---|---|---|---|
| Crude medical cost ratio (95% CI) | p | Adjusted medical cost ratio b (95% CI) | p | Crude day ratio (95% CI) | p | Adjusted day ratio c (95% CI) | p | |
| 75 years old | ||||||||
| Periodontitis | ||||||||
| No | 1 | 1 | 1 | 1 | ||||
| Moderate | 0.79 (0.25–2.48) | 0.684 | 0.69 (0.31–1.56) | 0.374 | 0.53 (0.13–2.18) | 0.526 | 0.40 (0.15–1.07) | 0.067 |
| Severe | 1.65 (0.47–5.78) | 0.436 | 1.41 (0.58–3.43) | 0.448 | 1.36 (0.28–6.65) | 0.703 | 1.24 (0.34–4.50) | 0.742 |
| 80 years old | ||||||||
| Periodontitis | ||||||||
| No | 1 | 1 | 1 | 1 | ||||
| Moderate | 3.01 (0.89–10.18) | 0.076 | 4.18 (1.71–10.20) | 0.002 | 7.67 (2.06–18.61) | 0.002 | 7.14 (2.18–23.42) | 0.001 |
| Severe | 5.07 (1.91–15.99) | 0.006 | 6.32 (2.21–18.06) | < 0.001 | 23.51 (4.37–126.41) | < 0.001 | 21.85 (5.94–80.35) | < 0.001 |
Abbreviation: CI, confidence interval.
Each age group was analysed separately.
Adjusted for sex, medical history of diabetes and hypertension, smoking status, economic status, body mass index, medical costs at baseline year and number of months since stroke.
Adjusted for sex, medical history of diabetes and hypertension, smoking status, economic status, body mass index and medical costs at baseline year.
4. Discussion
Participants with severe periodontitis had a significantly higher risk of stroke onset. Moreover, among patients with stroke, those with severe periodontitis incurred higher stroke‐related medical costs and had longer stroke‐related hospital stays.
A strength of our study is its longitudinal design, which provides more robust evidence than cross‐sectional studies. Furthermore, no previous studies had investigated the association between periodontitis and stroke onset, stroke‐related medical costs and length of hospital stay, emphasising the novel contribution of these findings. However, several limitations should be acknowledged. Information on social factors and lifestyle habits, other than smoking status, was limited. Stroke onset was determined using medical claims data, which may not precisely reflect the actual timing of the events. Because we used data from dental check‐ups aimed at maintaining the oral health of insured individuals, periodontitis was assessed using the CPI, a partial examination method. Consequently, the accuracy of periodontitis assessment may have been limited, and there is a risk that cases of periodontitis may have been overlooked or underestimated. In this case, the association between periodontitis and stroke may be biased towards the null. CPI, which is commonly used in public health initiatives to assess periodontal status at the population level, is not suitable for assessing disease severity or for diagnostic purposes. Consequently, discrepancies may exist between CPI scores and participants' actual periodontal conditions. The dental check‐ups were conducted across multiple clinics within a single prefecture. Although detailed manuals were used, examiner calibration was insufficient. Inter‐examiner variability may have led to misclassification of periodontitis. Because these errors are expected to occur independently of outcomes, they represent non‐differential misclassification and likely bias the association towards the null. In addition, participants were restricted to those with ≥ 20 teeth, to enable assessment of periodontitis. However, individuals with fewer teeth may have lost them due to periodontitis, and their exclusion may limit the external validity of this study. As dental check‐ups were conducted only for individuals who chose to attend nearby clinics, participants were more likely to be independent older adults with high health awareness. Changes in periodontal and overall health status over the 5‐year follow‐up period were not assessed. In addition, as the study population consisted of older adults aged 75 and 80 years living in a specific region, the findings may not be generalisable to other populations. Annual total medical costs were used as a proxy for general health status, which may not reflect the actual health condition. Hypertension was identified based on medical claims data; as blood pressure measurements were not obtained, individuals with undiagnosed hypertension who had not visited a medical institution may have been missed.
In this study of Japanese late‐stage older adults, individuals with severe periodontitis had a higher risk of stroke onset, suggesting that periodontitis may be a risk factor for stroke. Numerous studies have reported associations between periodontitis and cardiovascular disease or atherosclerosis; however, studies linking periodontal disease to stroke are limited. A 5‐year cohort study using medical claims data found that individuals diagnosed with periodontal disease or receiving periodontal treatment had a significantly higher HR for stroke onset (Hsu et al. 2022). Similarly, a meta‐analysis of cohort studies indicated that periodontitis is a risk factor for stroke (El Masri et al. 2024). Conversely, several studies have reported no significant association between periodontitis and stroke, leaving the relationship between the two unclear (Ma et al. 2023; Tiensripojamarn et al. 2021). Stroke incidence varies by region, with higher rates observed in low‐income countries in Asia and Africa, and by age, with rates increasing among individuals aged ≥ 65 years (GBD 2021 Stroke Risk Factor Collaborators 2024; Seo et al. 2014). Differences in study region, age distribution, follow‐up duration and the criteria used to classify periodontitis and stroke contribute to these inconsistencies.
DNA from oral bacteria, including periodontal pathogens, has been detected in atherosclerotic plaque deposits (Armingohar et al. 2014; Rafferty et al. 2011). In patients with periodontal disease, the tissue at the base of periodontal pockets is often ulcerated, facilitating the entry of subgingival bacteria into the bloodstream (Bosshardt 2018). Toothbrushing and periodontal treatment can also induce transient bacteraemia (Horliana et al. 2014; Tomás et al. 2012). Once in circulation, oral bacteria and their components may reach distant organs and trigger chronic inflammation. These microorganisms may contribute to stroke risk by promoting atherosclerotic plaque formation or destabilisation in blood vessels remote from the periodontal tissues.
Individuals with poor periodontal health often have low health awareness and engage in unhealthy lifestyle behaviours, resulting in a higher prevalence of conditions such as hypertension and diabetes mellitus (Sabine Elisabeth et al. 2025; Saito et al. 2025). Oral health behaviours, which influence periodontal status, may affect the overall health through lifestyle habits (Saito et al. 2025). Although our study was adjusted for age, lifestyle factors and medical history, residual confounding cannot be excluded. Periodontal disease and stroke share common risk factors, which may partly explain why participants with poor periodontal health exhibited higher stroke incidence, greater stroke‐related medical costs and longer hospital stays. Consequently, individuals with poor periodontal health are likely at elevated risk for stroke, and interventions aimed at reducing periodontal disease through targeted health education may help mitigate this risk.
Chronic inflammatory responses associated with periodontitis, including elevated systemic C‐reactive protein levels, may accelerate atherosclerosis and contribute to stroke onset (Badimon et al. 2018; Machado et al. 2021; Oh and Parikh 2022). In addition, periodontal disease increases the risk of metabolic disorders such as diabetes mellitus and dyslipidaemia, which may indirectly elevate stroke risk (Genco et al. 2020; Nepomuceno et al. 2017). Other potential mechanisms include dysbiosis induced by oral bacteria translocating from the oral cavity to the gut, which may influence stroke development (Zhong et al. 2024). These pathways suggest that periodontitis can increase the risk of stroke through multiple interrelated mechanisms.
In our study, analyses of participants with stroke suggested that those with severe periodontitis had higher stroke‐related medical costs and longer hospital stays compared to those without periodontitis, suggesting that periodontitis could influence stroke severity and recovery. Several studies have reported that individuals with good oral health incur lower medical costs (Iwasaki et al. 2017; Saito et al. 2019); however, no previous report has investigated the relationship between periodontitis and stroke‐related medical expenditures. Moreover, our study differs from previous studies by evaluating the associations of periodontitis with medical costs and length of hospital stay in stroke patients. Consistent with prior studies, individuals with poor oral health tended to incur higher medical costs and experience longer hospitalisations. Among patients with stroke, both stroke‐related medical costs and length of hospital stay were higher in 80‐year‐olds with severe periodontitis, whereas no significant association was observed in 75‐year‐olds. Compared to 75‐year‐olds, individuals aged 80 years experience greater age‐related declines in physical function. Consequently, older patients may have prolonged recovery from stroke, potentially resulting in longer hospital stays and higher medical costs. Although the analysis of medical costs and length of hospital stay for 80‐year‐olds in this study yielded statistically significant results, the small sample size and very wide CIs suggest that the precision of the estimates is limited. Therefore, these results are subject to uncertainty; caution is required when interpreting them, and further verification through large‐scale studies is necessary.
From the start of observation through 22 months, no significant associations were observed between periodontitis and stroke onset. However, from 23 months onwards, individuals with severe periodontitis had a significantly higher HR for stroke onset compared to those without periodontitis. Stroke onset is primarily driven by atherosclerosis, and the formation and progression of vascular plaques takes time. Therefore, it is plausible that periodontitis influenced stroke onset gradually through atherosclerotic progression. A nationwide, population‐based cohort study evaluating the association between changes in periodontal disease status and cardiovascular disease reported that individuals with persistent periodontal disease had a higher risk of developing cardiovascular disease compared to those without periodontal disease, those whose periodontal disease improved and those who newly developed periodontal disease (Park et al. 2025). In our study, however, the baseline dental check‐up did not necessarily reflect the onset of periodontitis in participants. The duration of periodontitis or the cumulative effects at baseline are unknown. In addition, factors not captured in our study, such as receipt of periodontal treatment during follow‐up or changes in overall health status, may have influenced the observed relationship.
5. Conclusion
Individuals with severe periodontitis had a higher risk of stroke. Moreover, among patients with stroke, those with severe periodontitis incurred higher stroke‐related medical costs and had longer stroke‐related hospital stays. These findings indicate that periodontitis may increase stroke risk and severity, thereby contributing to higher medical care costs.
Author Contributions
M.S. contributed to conception, design, analysis, interpretation and drafting and critically revising the manuscript. Y.S. contributed to conception, design, interpretation and drafting and critically revising the manuscript; T.N. and Y.I. contributed to conception, design, data acquisition and critically revising the manuscript. All authors gave their final approval and agree to be accountable for all aspects of the work.
Funding
This work was supported by the Japan Society for the Promotion of Science, 22K10353, 23K16239, 25K13330.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Table S1: Relationship between the incidence of stroke and independent variables during the follow‐up period according to Cox proportional hazards models with a cut‐off point of 1 year.
Table S2: Relationship between the incidence of stroke and independent variables during the follow‐up period according to Cox proportional hazards models with a cut‐off point of 2 years.
Table S3: Relationship between the incidence of stroke and independent variables during the follow‐up period according to Cox proportional hazards models with a cut‐off point of 3 years.
Table S4: Relationship between the incidence of stroke and independent variables during the follow‐up period according to Cox proportional hazards models with a cut‐off point of 4 years.
Supporting Information: Appendix A.
Supporting Information jcpe70157‐sup‐0002‐Supinfo
Contributor Information
Mizuki Saito, Email: saito@dpc.agu.ac.jp.
Yoshihiro Shimazaki, Email: shima@dpc.agu.ac.jp.
Data Availability Statement
The data that support the findings of this study are available from the Mie prefecture wide‐area union, but restrictions apply to the availability of these data, which were used under licence for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of the Mie prefecture wide‐area union.
References
- Armingohar, Z. , Jørgensen J. J., Kristoffersen A. K., and Abesha‐Belay E.. 2014. “Bacteria and Bacterial DNA in Atherosclerotic Plaque and Aneurysmal Wall Biopsies From Patients With and Without Periodontitis.” Journal of Oral Microbiology 6: 23408. 10.3402/jom.v6.23408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Asmat‐Abanto, A. S. , Espejo‐Carrera R. E., Honores‐Solano T. M., Del Castillo‐Huertas O. M., Caballero‐Alvarado J. A., and Minchón‐Medina C. A.. 2025. “Is Periodontitis a Risk Factor for Ischemic Stroke?: Systematic Review and Meta‐Analysis.” Journal of Clinical and Experimental Dentistry 17, no. 3: e329–e340. 10.4317/jced.62538. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Badimon, L. , Peña E., Arderiu G., et al. 2018. “C‐Reactive Protein in Atherothrombosis and Angiogenesis.” Frontiers in Immunology 9: 430. 10.3389/fimmu.2018.00430. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bell, S. , Gibson J. T., Harshfield E. L., and Markus H. S.. 2020. “Is Periodontitis a Risk Factor for Ischaemic Stroke, Coronary Artery Disease and Subclinical Atherosclerosis? A Mendelian Randomization Study.” Atherosclerosis 313: 111–117. 10.1016/j.atherosclerosis.2020.09.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Bosshardt, D. D. 2018. “The Periodontal Pocket: Pathogenesis, Histopathology and Consequences.” Periodontology 2000 76, no. 1: 43–50. 10.1111/prd.12153. [DOI] [PubMed] [Google Scholar]
- Dietrich, T. , Sharma P., Walter C., Weston P., and Beck J.. 2013. “The Epidemiological Evidence Behind the Association Between Periodontitis and Incident Atherosclerotic Cardiovascular Disease.” Journal of Clinical Periodontology 40, no. Suppl 14: S70–S84. 10.1111/jcpe.12062. [DOI] [PubMed] [Google Scholar]
- El Masri, J. , Al Malak A., El Masri D., et al. 2024. “The Association Between Periodontitis, Gingivitis, Tooth Loss and Stroke: An Umbrella Study With Meta‐Analysis.” Brain Sciences 15, no. 1: 10. 10.3390/brainsci15010010. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fagundes, N. C. F. , Almeida A., Vilhena K. F. B., Magno M. B., Maia L. C., and Lima R. R.. 2019. “Periodontitis as A Risk Factor for Stroke: A Systematic Review and Meta‐Analysis.” Vascular Health and Risk Management 15: 519–532. 10.2147/vhrm.S204097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- GBD 2021 Stroke Risk Factor Collaborators . 2024. “Global, Regional, and National Burden of Stroke and Its Risk Factors, 1990‐2021: A Systematic Analysis for the Global Burden of Disease Study 2021.” Lancet Neurology 23, no. 10: 973–1003. 10.1016/s1474-4422(24)00369-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Genco, R. J. , Graziani F., and Hasturk H.. 2020. “Effects of Periodontal Disease on Glycemic Control, Complications, and Incidence of Diabetes Mellitus.” Periodontology 2000 83, no. 1: 59–65. 10.1111/prd.12271. [DOI] [PubMed] [Google Scholar]
- Horliana, A. C. , Chambrone L., Foz A. M., et al. 2014. “Dissemination of Periodontal Pathogens in the Bloodstream After Periodontal Procedures: A Systematic Review.” PLoS One 9, no. 5: e98271. 10.1371/journal.pone.0098271. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hsu, P. W. , Shen Y. W., Syam S., et al. 2022. “Patients With Periodontitis Are at a Higher Risk of Stroke: A Taiwanese Cohort Study.” Journal of the Chinese Medical Association 85, no. 10: 1006–1010. 10.1097/jcma.0000000000000797. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Iwasaki, M. , Sato M., Yoshihara A., Ansai T., and Miyazaki H.. 2017. “Association Between Tooth Loss and Medical Costs Related to Stroke in Healthy Older Adults Aged Over 75 Years in Japan.” Geriatrics & Gerontology International 17, no. 2: 202–210. 10.1111/ggi.12687. [DOI] [PubMed] [Google Scholar]
- Lafon, A. , Pereira B., Dufour T., et al. 2014. “Periodontal Disease and Stroke: A Meta‐Analysis of Cohort Studies.” European Journal of Neurology 21, no. 9: 1155–1161. 10.1111/ene.12415. [DOI] [PubMed] [Google Scholar]
- Leira, Y. , Seoane J., Blanco M., et al. 2017. “Association Between Periodontitis and Ischemic Stroke: A Systematic Review and Meta‐Analysis.” European Journal of Epidemiology 32, no. 1: 43–53. 10.1007/s10654-016-0170-6. [DOI] [PubMed] [Google Scholar]
- Ma, C. , Wu M., Gao J., et al. 2023. “Periodontitis and Stroke: A Mendelian Randomization Study.” Brain and Behavior: A Cognitive Neuroscience Perspective 13, no. 2: e2888. 10.1002/brb3.2888. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Machado, V. , Botelho J., Escalda C., et al. 2021. “Serum C‐Reactive Protein and Periodontitis: A Systematic Review and Meta‐Analysis.” Frontiers in Immunology 12: 706432. 10.3389/fimmu.2021.706432. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Meng, X. , and Chen X.. 2025. “Periodontitis and Risk of Stroke: A Systematic Review and Meta‐Analysis of Observational Studies.” Frontiers in Neurology 16: 1700946. 10.3389/fneur.2025.1700946. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ministry of Health, Labour and Welfare . 2022. “Comprehensive Survey of Living Conditions.” https://www.mhlw.go.jp/toukei/saikin/hw/k‐tyosa/k‐tyosa22/dl/05.pdf.
- Ministry of Health, Labour and Welfare . 2024. “Vital Statistics.” https://www.mhlw.go.jp/toukei/saikin/hw/jinkou/kakutei24/dl/11_h9.pdf.
- Nepomuceno, R. , Pigossi S. C., Finoti L. S., et al. 2017. “Serum Lipid Levels in Patients With Periodontal Disease: A Meta‐Analysis and Meta‐Regression.” Journal of Clinical Periodontology 44, no. 12: 1192–1207. 10.1111/jcpe.12792. [DOI] [PubMed] [Google Scholar]
- Oh, S. E. , and Parikh N. S.. 2022. “Recent Advances in the Impact of Infection and Inflammation on Stroke Risk and Outcomes.” Current Neurology and Neuroscience Reports 22, no. 3: 161–170. 10.1007/s11910-022-01179-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Park, J. H. , Leem G. H., Kim J. W., and Song T. J.. 2025. “Persisting Chronic Periodontal Disease as a Risk Factor for Cardiovascular Disease: A Nationwide Population‐Based Cohort Study.” Journal of Clinical Periodontology 52, no. 3: 375–386. 10.1111/jcpe.14107. [DOI] [PubMed] [Google Scholar]
- Rafferty, B. , Jönsson D., Kalachikov S., et al. 2011. “Impact of Monocytic Cells on Recovery of Uncultivable Bacteria From Atherosclerotic Lesions.” Journal of Internal Medicine 270, no. 3: 273–280. 10.1111/j.1365-2796.2011.02373.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sabine Elisabeth, G. , Yuxi Z., Jiawen Y., Lei W., Sabine R., and Joerg M.. 2025. “Systemic, Lifestyle and Environmental Modifying Factors in the Pathogenesis of Periodontitis.” Journal of Periodontal Research. 10.1111/jre.70003. [DOI] [PubMed] [Google Scholar]
- Saito, M. , Shimazaki Y., Nonoyama T., and Ohsugi K.. 2021. “Association Between Oral Health and the Medical Costs of Dementia: A Longitudinal Study of Older Japanese.” American Journal of Alzheimer's Disease and Other Dementias 36: 1533317521996142. 10.1177/1533317521996142. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Saito, M. , Shimazaki Y., Nonoyama T., and Tadokoro Y.. 2019. “Associations of Number of Teeth With Medical Costs and Hospitalization Duration in an Older Japanese Population.” Geriatrics & Gerontology International 19, no. 4: 335–341. 10.1111/ggi.13622. [DOI] [PubMed] [Google Scholar]
- Saito, M. , Shimazaki Y., Yoshii S., and Kojima T.. 2025. “Oral Health Behaviours and Metabolic Syndrome: Mediation Effect of Lifestyle Habits.” BMC Oral Health 25, no. 1: 510. 10.1186/s12903-025-05860-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Schenkein, H. A. , Papapanou P. N., Genco R., and Sanz M.. 2020. “Mechanisms Underlying the Association Between Periodontitis and Atherosclerotic Disease.” Periodontology 2000 83, no. 1: 90–106. 10.1111/prd.12304. [DOI] [PubMed] [Google Scholar]
- Seo, S. R. , Kim S. Y., Lee S. Y., et al. 2014. “The Incidence of Stroke by Socioeconomic Status, Age, Sex, and Stroke Subtype: A Nationwide Study in Korea.” Journal of Preventive Medicine and Public Health. Yebang Uihakhoe Chi 47, no. 2: 104–112. 10.3961/jpmph.2014.47.2.104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tiensripojamarn, N. , Lertpimonchai A., Tavedhikul K., et al. 2021. “Periodontitis Is Associated With Cardiovascular Diseases: A 13‐Year Study.” Journal of Clinical Periodontology 48, no. 3: 348–356. 10.1111/jcpe.13418. [DOI] [PubMed] [Google Scholar]
- Tomás, I. , Diz P., Tobías A., Scully C., and Donos N.. 2012. “Periodontal Health Status and Bacteraemia From Daily Oral Activities: Systematic Review/Meta‐Analysis.” Journal of Clinical Periodontology 39, no. 3: 213–228. 10.1111/j.1600-051X.2011.01784.x. [DOI] [PubMed] [Google Scholar]
- Toyoda, K. , Inoue M., and Koga M.. 2019. “Small but Steady Steps in Stroke Medicine in Japan.” Journal of the American Heart Association 8, no. 16: e013306. 10.1161/jaha.119.013306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Villoria, G. E. M. , Fischer R. G., Tinoco E. M. B., Meyle J., and Loos B. G.. 2024. “Periodontal Disease: A Systemic Condition.” Periodontology 2000 96, no. 1: 7–19. 10.1111/prd.12616. [DOI] [PMC free article] [PubMed] [Google Scholar]
- World Health Organization . 1997. Oral Health Surveys, Basic Methods. 4th ed. World Health Organization. [Google Scholar]
- Zhong, Y. , Kang X., Bai X., et al. 2024. “The Oral‐Gut‐Brain Axis: The Influence of Microbes as a Link of Periodontitis With Ischemic Stroke.” CNS Neuroscience & Therapeutics 30, no. 12: e70152. 10.1111/cns.70152. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Table S1: Relationship between the incidence of stroke and independent variables during the follow‐up period according to Cox proportional hazards models with a cut‐off point of 1 year.
Table S2: Relationship between the incidence of stroke and independent variables during the follow‐up period according to Cox proportional hazards models with a cut‐off point of 2 years.
Table S3: Relationship between the incidence of stroke and independent variables during the follow‐up period according to Cox proportional hazards models with a cut‐off point of 3 years.
Table S4: Relationship between the incidence of stroke and independent variables during the follow‐up period according to Cox proportional hazards models with a cut‐off point of 4 years.
Supporting Information: Appendix A.
Supporting Information jcpe70157‐sup‐0002‐Supinfo
Data Availability Statement
The data that support the findings of this study are available from the Mie prefecture wide‐area union, but restrictions apply to the availability of these data, which were used under licence for the current study, and so are not publicly available. Data are however available from the authors upon reasonable request and with permission of the Mie prefecture wide‐area union.