The Impact of Periodontal Health Interventions on the Risk of Ischemic Stroke: A Systematic Review and Meta‐Analysis

ABSTRACT

Aim

To systematically review the evidence on the relationship between periodontal health interventions and the risk of first‐ever ischemic stroke.

Materials and Methods

Search strategy included electronic and hand searching to January 2025. MEDLINE (OVID), EMBASE (OVID), SCOPUS, LILACS, Web of Science (WOS) and Cochrane Central Register of Controlled Trials (CENTRAL) were searched. Eligible studies included clinical trials and observational designs reporting on any periodontal health intervention and ischemic stroke. Screening and data extraction were performed independently and in duplicate. Meta‐analyses were conducted when appropriate.

Results

Seven studies involving 1,886,947 participants met the inclusion criteria from an initial set of 634 records. In a random‐effects meta‐analysis, periodontal interventions were associated with a lower risk of ischemic stroke (HR: 0.88 [0.81–0.95]). Tooth scaling, in particular, was associated with a reduced risk (HR: 0.84 [0.76–0.92]).

Conclusion

Periodontal health interventions—especially tooth scaling—appear to be associated with a lower risk of ischemic stroke. These findings suggest a potential role for periodontal care in stroke prevention.

Trial Registration

PROSPERO number: CRD42023450471

1. Introduction

Ischemic stroke is a major cause of morbidity and mortality worldwide, particularly in developed countries [1]. Its high incidence and substantial health impact have motivated extensive research into its risk factors and underlying mechanisms.

Periodontal disease has emerged as a topic of growing interest in this context. Recent epidemiological studies and meta‐analyses have reported an association between periodontitis and the incidence of ischemic stroke [2, 3].

Biological plausibility supports this relationship. Periodontal pathogens and their by‐products can enter the systemic circulation, promoting a state of chronic low‐grade inflammation [4, 5]. This inflammatory response contributes to endothelial dysfunction and to the development and destabilization of atherosclerotic plaques, processes that are particularly relevant for atherothrombotic ischemic stroke [6].

Given that periodontitis is one of the most prevalent chronic inflammatory diseases worldwide [7, 8], it is clinically relevant to explore whether periodontal health interventions may reduce the risk of ischemic stroke. These interventions include both preventive measures, such as regular dental visits, oral hygiene instructions, toothbrushing and tooth scaling, as well as therapeutic approaches aimed at controlling established disease, including scaling and root planing and periodontal flap surgery [9]. By improving periodontal health, these strategies may attenuate systemic inflammation and potentially lower cerebrovascular risk.

Although several studies have examined this association, their findings remain heterogeneous regarding populations, types of interventions and outcomes assessed. To date, no review has systematically evaluated this evidence or explored the sources of variability across studies. Therefore, the aim of the present systematic review and meta‐analysis was to assess whether periodontal health interventions are associated with a reduced risk of first‐ever ischemic stroke and to evaluate the consistency and clinical relevance of the available evidence.

2. Materials and Methods

The research protocol for this systematic review was registered in PROSPERO on 23 August 2023 with ID number CRD42023450471, following the PRISMA 2020 guidelines [10].

2.1. Eligibility: Inclusion and Exclusion Criteria

The research questions were defined using the PICOS outline (Table 1).

TABLE 1.

PICOS outline.

Population	Intervention	Comparison	Outcome	Study design
Individuals aged more than 18 years old with or without periodontitis.	Exposure to any periodontal health intervention whether invasive (scaling and root planning, periodontal flap surgery, tooth extraction) or non‐invasive (dental visits, toothbrushing, tooth scaling).	No treatment and/or a different treatment to the one that is being studied.	Risk of ischemic stroke.	Randomized and non‐randomized clinical trials, and observational studies (cohort, case–control and cross‐sectional studies).
What is the effect of any periodontal health intervention on the risk of suffering from ischemic stroke compared to non‐intervention or controlled treatment in adults?

Studies were eligible for inclusion in the review if they were randomized or non‐randomized clinical trials, or observational studies (case–control studies, cohort studies or cross‐sectional studies) reporting on the risk of incident/prevalent ischemic stroke in individuals over 18 years of age undergoing any periodontal health intervention. Only studies providing a clear description of the treatment were included.

In this way, systematic reviews with or without meta‐analysis, consensus documents, informative documents, summaries, in vitro studies and studies on animal models were not eligible for inclusion in our review. Furthermore, those studies about global cardiovascular disease and studies about haemorrhagic strokes or that do not distinguish this from ischemic stroke were also not included.

The date of publication, country and language were not considered when selecting the studies.

2.2. Treatment Definitions

All periodontal health interventions were considered eligible, regardless of their complexity or preventive or therapeutic intent, provided that they were clearly described and compared with non‐intervention or an alternative treatment. This broad inclusion was intended to capture the full spectrum of periodontal care and to evaluate its global association with ischemic stroke risk. This approach resulted in substantial variability in how oral hygiene and periodontal interventions were defined and reported across studies. To address this heterogeneity, we retained each study's original exposure definitions and grouped the interventions into six broader categories according to their complexity:

1. Dental visits: frequency of visits to the dentist.

2. Toothbrushing: frequency of toothbrushing with either manual or electric toothbrushes.

3. Tooth scaling: tooth cleaning to remove supragingival plaque.

4. Scaling and root planing: subgingival debridement to remove subgingival plaque.

5. Periodontal surgery: any periodontal surgical procedure including open flap debridement (i.e., access periodontal surgery), pocket elimination/reduction periodontal surgery (i.e., resective periodontal surgery) or regenerative periodontal surgery.

6. Tooth extraction: extraction of hopeless teeth (i.e., irrational to treat).

However, the original definitions and reporting of each intervention were preserved for descriptive analysis, and detailed variations in exposure assessment across individual studies are presented in the Results section.

2.3. Information Sources and Searches

To carry out this systematic review, a set of six different databases was searched up to January 2025, including MEDLINE (OVID), EMBASE (OVID), SCOPUS, LILACS, Web of Science (WOS) and Cochrane Central Register of Controlled Trials (CENTRAL), using the detailed search strategy that is shown below: (‘dental treatment’ OR ‘periodontal treatment’ OR ‘dental prophylaxis’ OR ‘dental scaling’ OR ‘tooth scaling’) AND (‘ischemic stroke’ OR ‘stroke’ OR ‘cerebrovascular disease’).

In addition, each systematic review, consensus document, or article registration that was found with the search strategy, although it was not included, was reviewed in order to collect those target articles that escaped the initial search.

2.4. Study Selection and Data Extraction

Titles and abstracts (when available) were independently screened by two reviewers for eligibility based on inclusion/exclusion criteria (JM, YL). Full‐text reports were obtained and assessed independently and in duplicate for studies appearing to meet the inclusion criteria or with insufficient information in the title or abstract to confirm eligibility for inclusion. Any disagreement was resolved by discussion with a third reviewer (RL). In terms of study exclusion and inclusion, the scores for the 2 reviewers' agreement were 90% at the first step for screening titles and abstracts and 100% at the second step for screening full‐text articles.

Data on study characteristics such as type of study, country, population, sample size, definitions, interventions, adjusted covariates and outcomes were recorded in evidence tables to provide an overview of the included studies and available data. All data in the Excel spreadsheets were reviewed to consider appropriateness for meta‐analysis.

2.5. Risk of Bias Assessment

Quality assessment of included studies was undertaken independently and in duplicate by two reviewers (JM, YL) using the Newcastle‐Ottawa Quality Scale (NOS) [11], designed for cohort studies. Disagreements were resolved by discussion with a third reviewer (RL).

2.6. Quality of Evidence Assessment

The analysis of the quality and certainty of the evidence, which defines the degree of confidence in the obtained results, was carried out independently and in duplicate by two reviewers (JM, YL) using the GRADE System [12, 13]. Disagreements were resolved by discussion and if necessary, a third reviewer was consulted (RL).

2.7. Data Analysis and Synthesis

Data was shown in evidence tables, grouped according to study design. Then, a descriptive analysis was performed to determine the quantity of data, checking further for study variations in terms of study characteristics and results. This facilitated the assessment of whether further synthesis methods were appropriate. If quantitative methods were possible, pooled hazard ratios (HR) and 95% confidence intervals were calculated for extracted data. Because clinical heterogeneity was anticipated a priori, random‐effects models were used for all meta‐analyses. Statistical heterogeneity was also quantified using Cochran's Q test [14] and the I ² statistic [15] to describe the degree of variability among studies.

Potential publication bias was examined using a funnel plot and Egger's test [16]. For meta‐analyses showing nominal statistical significance, we calculated Rosenberg's FSN to determine the quantity of future studies with average null effect and average weight required to yield a summary estimate that was not statistically significant [17]. For our analysis, we employed the Meta‐Essentials software packages to handle both binary and continuous measures.

3. Results

3.1. Study Selection

The electronic search identified 634 records, and one additional article was retrieved through citation screening. After removal of duplicates, 325 records were screened by title and abstract. Thirteen full‐text articles were assessed for eligibility. Of these, six studies were excluded for the following reasons: inclusion of stroke as part of a composite cardiovascular outcome (n = 3) [18, 19, 20], inclusion of hemorrhagic stroke (n = 1) [21], provision of indirect evidence (n = 1) [22], and assessment of stroke recurrence rather than incident stroke (n = 1) [23]. Consequently, seven studies met the inclusion criteria and were included in the qualitative synthesis (Figure 1) [24, 25, 26, 27, 28, 29, 30]. Six of these studies were eligible for quantitative analysis [24, 25, 26, 28, 29, 30].

FIGURE 1.

PRISMA flow chart.

3.2. Study Characteristics—Description of Studies

Seven cohort studies published between 2012 and 2025 were included [24, 25, 26, 27, 28, 29, 30]. An evidence table summarizing all seven studies was generated according to study design and presented in chronological order (Table 2).

TABLE 2.

Main characteristics of included studies.

Study	Design	State	Source	Sample size	PD definition	IS definition	Intervention	Follow up	Adjusted covariates	Results: adjusted HR (95% CI)
Chen et al. (2012)	Prospective cohort	1,000,000 patients from LHID 2000 (NHIRD) collected in year 2000	21,876 patients (10,887 treated) (10,989 untreated)	Not reported	Modified ICD‐9 Criteria (1 time in hospital and 2 times in outpatient care)	Tooth Scaling	7 years	Age, Sex, HBP, DM, Dyslipidemia, Arrhythmias, CKD.	Tooth Scaling: 0.85 [0.78–0.93] p < 0.001 < 65: 0.77 [0.57–1.03] ≥ 65: 0.67 [0.51–0.87] p < 0.05 Man: 0.75 [0.59–0.96] p < 0.05 Women: 0.65 [0.47–0.91] p < 0.05 HBP yes: 0.65 [0.50–0.84] p < 0.05 HBP no: 0.81 [0.59–1.12] Diabetes Mellitus yes: 0.62 [0.45–0.86] p < 0.05 Diabetes Mellitus no: 0.77 [0.60–0.99] p < 0.05 Dyslipidemia yes: 0.65 [0.43–0.98] p < 0.05 Dyslipidemia no: 0.74 [0.59–0.93] p < 0.05 Coronary Heart Disease yes: 0.63 [0.43–0.94] p < 0.05 Coronary Heart Disease no: 0.74 [0.59–0.94] p < 0.05 ≥ 1 Tooth Scaling every 2 years: 0.81 [0.73–0.92] p < 0.001 < 1 Tooth Scaling every 2 years: 0.92 [0.83–1.02] p < 0.001
Lee et al. (2013)	Retrospective cohort	1,000,000 patients from LHID 2000 (NHIRD) collected between 2000 and 2010	719,436 patients (208,674 without PD) (510,762 with PD)	Modified ICD‐9 Criteria	Modified ICD‐9 Criteria	Tooth Scaling Intensive treatment (scaling and root planing, periodontal flap or tooth extraction)	8.32 years	Age, Sex, HBP, DM, Dyslipidemia, AF, Atherosclerosis, CKD.	Untreated: 1.15 [1.07–1.24] p < 0.05. By age: 20–44: 2.17 [1.64–2.87] p < 0.05 45–64: 1.19 [1.05–1.35] p < 0.05 ≥ 65: 1.13 [1.03–1.25] p < 0.05 Tooth Scaling: 0.78 [0.75–0.81] p < 0.05. By age: 20–44: 0.74 [0.64–0.86] p < 0.05 45–64: 0.68 [0.64–0.73] p < 0.05 ≥ 65: 0.99 [0.94–1.05] Intensive treatment: 0.95 [0.91–0.99] p < 0.05. By age: 20–44: 1.35 [1.12–1.62] p < 0.05 45–64: 0.89 [0.83–0.96] p < 0.05 ≥ 65: 1.02 [0.96–1.08]
Sen et al. (2018)	Prospective cohort	USA	Patients from ARICs 4th visit (1996–1998)	10,362 patients (6670 regular care) (3692 episodic care)	Not reported	ICD‐10 Criteria	Regular dental care Episodic dental care	15 years	Age, Sex, HBP, DM, BMI, LDL‐c, Smoking, Education, Race.	Regular dental care: 0.77 [0.63–0.94] p < 0.05
Huang et al. (2018)	Retrospective cohort	CKD‐Haemodialysis patients from RCID (NHIRD) collected between 1998 and 2011	11,318 patients (4149 treated) (7169 untreated) 7226 matched (3613 treated) (3613 untreated)	Modified ICD‐9 Criteria	Modified ICD‐9 Criteria	Intensive treatment (scaling and root planing or periodontal flap)	4.97 years	Age, Sex, Income, Urbanization level, Charlson Index, HBP, DM, Dyslipidemia, COPD, Peripheral vascular disease, amputation, Antiplatelets, ACEI, ARB, Statins, Calcium antagonists and β‐Blockers.	Intensive treatment: 0.78 [0.68–0.90] p < 0.001. ≤ 49: 0.58 [0.44–0.78] p < 0.001 50–64: 0.75 [0.62–0.90] p < 0.01 ≥ 65: 0.62 [0.51–0.76] p < 0.001 Man: 0.71 [0.60–0.85] p < 0.001 Women: 0.65 [0.55–0.77] p < 0.001 Comorbidity yes: 0.67 [0.59–0.76] p < 0.001 Comorbidity no: 0.79 [0.51–1.23]
Lin et al. (2019)	Retrospective cohort	1,000,000 patients from LHID 2005 (NHIRD) enrolled and collected in year 2005	161,923 patients (74,516 gingivitis) (87,407 periodontitis)	Modified ICD‐9 Criteria (≥ 3 times in 6 months)	Modified ICD‐9 Criteria	Tooth Scaling Intensive treatment (scaling and root planning or periodontal flap) Tooth extraction Combinations	7.74 years	Age, Sex, HBP, DM‐II, COPD, Depression, Anxiety, Insomnia, Sleep Disorders, Traumatic Brain Injury and CKD.	Against no treatment: Tooth Scaling: 0.77 [0.66–0.89] p = 0.0005 Intensive treatment: 0.80 [0.69–0.93] p = 0.0037 Tooth extraction: 1.17 [1.04–1.33] p = 0.012 Against tooth scaling: Untreated: 1.35 [1.13–1.63] p = 0.0013 Intensive treatment: 0.62 [0.31–1.25] Tooth extraction: 1.56 [1.19–2.03] p = 0.0012 Tooth scaling and Intensive treatment: 0.82 [0.69–0.98] p = 0.0256 Tooth Scaling and tooth extraction: 1.16 [1.18–1.35] p = 0.0516 Intensive treatment and tooth extraction: 1.40 [0.52–3.75] Tooth scaling, intensive treatment and tooth extraction: 0.99 [0.73–1.35]
Chang et al. (2021)	Retrospective cohort	South Korea	Patients aged 40–79 from NHAC (KNHID)	206,602 patients (163,670 without PD) (42,932 with PD)	Korean revised ICD‐10 Criteria	ICD‐10 Criteria (2 visits in a year and imaging test)	Toothbrushing Dental visits Tooth Scaling	10.4 years	Age, Sex, Income, Physical Activity, Alcoholism, Smoking, BMI, HBP, DM, Dyslipidemia, CKD, Cancer, Systolic Pressure, Blood Glucose, Total Cholesterol, GOT, GPT, GGT, Proteinuria, Periodontal Disease, Visits to the dentist, Dental cleaning, Frequency of toothbrushing, Missing teeth and Number of cavities.	Dental Visits: 0.91 [0.86–0.96] p < 0.001 Tooth scaling: 0.94 [0.88–1.01] p = 0.083 Toothbrushing 2 times a day: 0.89 [0.84–0.95] p < 0.001 Toothbrushing ≥ 3 times a day: 0.81 [0.76–0.87] p < 0.001
Wagner et al. (2023)	Retrospective cohort	Denmark	Patients aged 30–90 from DHNSR with 1st MI between 2000 and 2015	103,949 patients (10,049 treated) (50,569 no dental visits) (43,331 healthy)	2010 Dental Care Labour Agreement Criteria	ICD‐10 Criteria	Scaling and root planing	3 years	Age, Sex, Year, HBP, DM, Chronic Ischemic Heart Disease, COPD/Smoking/Lung Cancer/Substitute Drug, CKD, Liver Disease, Peripheral Arterial Disease, AF, Educational status and Socioeconomic status.	Against healthy controls in dental visits (OR): Treated periodontitis in dental visits: 1.00 [0.85–1.17] Treated periodontitis in 1 tooth in dental visits: 1.06 [0.86–1.30] Treated periodontitis ≥ 2 teeth in dental visits: 0.94 [0.75–1.14] Treated periodontitis in 1 dental visit: 0.93 [0.75–1.30] Treated periodontitis ≥ 2 dental visits: 1.10 [0.88–1.38] Without visits to the dentist: 1.01 [0.92–1.10]

Abbreviations: DNHSR, Danish National Health Service Registry; IS, ischemic stroke; KNHID, Korean National Health Insurance Database; LHID, Longitudinal Health Insurance Database; MI, myocardial infarction; NHAC, National Health Assessment Cohort; NHIRD, National Health Insurance Research Database; PD, periodontal disease; RCID, Catastrophic Illness Registry Database.

Two studies had a prospective design and five were retrospective. Together, they included 1,886,947 individuals aged over 18 years, with 930,314 participants in the intervention groups and 956,633 in the control groups. The studies were conducted in three different countries and evaluated populations with heterogeneous characteristics.

Most studies assessed individuals from the general community [24, 25, 27, 28, 29], whereas two focused on specific high‐risk populations, including patients undergoing haemodialysis [26] and patients with previous myocardial infarction [30].

Three studies were conducted using the same national database in Taiwan [25, 28, 29]. One study used the 2005 cohort [25], while two studies used the 2000 cohort [28, 29]. Among the two studies based on the 2000 cohort, one included 21,000 individuals who received tooth scaling [29], and the other included 720,000 patients diagnosed with periodontal disease between 2000 and 2010 [28]. Therefore, there is a slight possibility of overlap among participants who both received tooth scaling and were diagnosed with periodontal disease; however, the majority of participants in each study represent distinct populations. Given that the studies focus on different aspects of oral health interventions, both were included in the meta‐analysis. Any potential overlap is expected to have minimal impact on the pooled estimates, and this was accounted for by using random‐effects models.

Follow‐up duration varied across studies, ranging from 3 to 15 years (Table 2), which was sufficient in all cases to capture incident ischemic stroke.

A wide range of periodontal health interventions was evaluated. These ranged from less complex measures, such as dental visits, toothbrushing and tooth scaling, to more intensive procedures, including scaling and root planing, periodontal flap surgery and tooth extraction. Some studies also examined combinations of these interventions [25].

3.3. Risk of Bias of Selected Studies

Overall, the included studies were judged to be of good methodological quality and at low risk of bias (Table 3). The main limitation identified was insufficient adjustment for smoking status in several studies. In addition, lower representativeness of the exposed cohort was observed in studies conducted in specific populations [26, 30].

TABLE 3.

Quality assessment for observational studies using Newcastle‐Ottawa scale.

Study	Selection				Comparability		Results			Total score (9/9*)	Risk of bias
	Representativeness of the exposed cohort	Selection of the non‐exposed cohort	Ascertainment of exposure	Outcome was not present at the start of the study	Comparability of cohorts for age, sex and HBP	Comparability of cohorts for smoking	Outcome evaluation	Follow up	Assumable loses
Chen et al. (2012)	*	*	*	*	*	0	*	*	*	8*	LOW
Lee et al. (2013)	*	*	*	*	*	0	*	*	*	8*	LOW
Sen et al. (2018)	*	*	*	*	*	*	*	*	*	9*	LOW
Huang et al. (2018)	0	*	*	*	*	0	*	*	*	7*	LOW
Lin et al. (2019)	*	*	*	*	*	0	*	*	*	8*	LOW
Chang et al. (2021)	*	*	0	*	*	*	*	*	*	8*	LOW
Wagner et al. (2023)	0	*	*	*	*	0	*	*	*	7*	LOW

*Indicates compliance with the corresponding item.

3.4. Descriptive Synthesis

3.4.1. Dental Visits

Definitions of dental visits varied across the included studies. Exposure was defined either as the presence versus absence of at least one dental visit within a given year, regardless of indication [24], or according to the purpose of the visit, distinguishing routine or preventive care from symptom‐driven or episodic attendance [27]. Details regarding provider type and specific procedures performed during visits were inconsistently reported.

Despite these variations, regular dental care was consistently associated with a reduced risk of incident ischemic stroke. Compared with episodic or symptom‐driven attendance, routine or preventive dental visits were associated with a lower risk (HR: 0.77 [0.63–0.94]) [27]. A similar protective association was observed when regular dental care was compared with no dental visits (HR: 0.91 [0.86–0.96]) [24].

3.4.2. Toothbrushing

Toothbrushing was assessed in one study as a self‐reported oral hygiene behaviour and categorized according to daily frequency [24]. Details regarding the characteristics of toothbrushing were not reported.

Higher toothbrushing frequency was associated with a lower risk of ischemic stroke. Compared with individuals who brushed once daily or less, those who brushed twice daily showed a reduced risk (HR: 0.89 [0.84–0.95]), while brushing three or more times per day was associated with a greater reduction (HR: 0.81 [0.76–0.87]) [24].

3.4.3. Tooth Scaling

Tooth scaling was evaluated in four of the included studies [24, 25, 28, 29]. Among these, one study explicitly distinguished between full‐mouth and localized tooth scaling, and both modalities were included under the tooth scaling category in the present review [29], whereas the remaining studies did not specify the treatment extent. Details regarding the provider of care were inconsistently reported. In studies conducted in Taiwan [25, 28, 29], scaling procedures were performed by dentists [31], whereas in South Korea [24] the procedure may have been carried out by either dentists or dental hygienists, although this was not explicitly specified in all cases [32, 33, 34].

Overall, the association between tooth scaling and ischemic stroke varied by population. In the general population, findings were inconsistent: one study reported a significant protective effect (HR: 0.85 [0.78–0.93]) [29], particularly when scaling was performed at least once every 2 years (HR: 0.81 [0.73–0.92]) [29], whereas another study found no statistically significant association (HR: 0.94 [0.88–1.01]) [24]. In contrast, among individuals with periodontal disease, tooth scaling was consistently associated with a reduced risk of ischemic stroke. Treated patients had a lower risk compared with untreated patients with periodontal disease [25] and also compared with untreated individuals without periodontal disease (HR: 0.78 [0.75–0.81]) [28].

Subgroup analyses showed that the protective effect of tooth scaling was maintained in individuals over 65 years of age in the general population [29], and in individuals under 65 years of age with periodontal disease [28]. The association was consistent across sex and most comorbidity subgroups, except in individuals without hypertension [29].

3.4.4. Intensive Periodontal Treatment

Intensive periodontal treatment was operationally defined in this review as the set of procedures including scaling and root planing, periodontal surgery (e.g., flap surgery) and tooth extraction. In the original studies, these procedures were sometimes analysed individually and sometimes jointly under composite terms such as ‘intensive periodontal treatment.’ Regardless of the original reporting, all three procedures were grouped under this category for analytical purposes, while the individual definitions were preserved for descriptive comparison.

Details regarding the delivery of intensive periodontal treatment were limited. In the studies conducted in Taiwan [25, 26], procedures were performed by dentists in accordance with national regulations, whereas in other settings, such as Denmark [30], scaling and root planing could be performed by either dentists or dental hygienists under national legislation [35], although the provider was not specified at the individual level. Information on the number of sessions, treatment protocols, use of adjunctive therapies and other procedural characteristics was not consistently reported across studies.

Most studies reported a protective association between intensive periodontal treatment and ischemic stroke. Treated patients with periodontal disease had a lower risk compared with untreated patients with periodontal disease (HR: 0.80 [0.69–0.93]) [25], and also compared with untreated individuals without periodontal disease (HR: 0.95 [0.91–0.99]) [28]. However, tooth extraction, when analysed as an isolated intervention, was associated with an increased risk of ischemic stroke (HR: 1.17 [1.04–1.33]) [25].

Age‐stratified analyses showed heterogeneous results. A protective effect was observed in individuals aged 45–64 years (HR: 0.89 [0.83–0.96]) [28]. In contrast, no significant association was found in individuals over 65 years of age (HR: 1.02 [0.96–1.08]) [28]. Among younger adults aged 20–44 years, intensive treatment was associated with an increased risk of ischemic stroke (HR: 1.35 [1.12–1.62]) [28].

In special populations, results differed by clinical context. In patients with a previous myocardial infarction, intensive periodontal treatment was not associated with a reduced risk of ischemic stroke (HR: 1.00 [0.85–1.17]) [30]. In contrast, a protective effect was observed among patients undergoing haemodialysis (HR: 0.78 [0.68–0.90]) [26], which was consistent across age and sex subgroups and in patients with comorbidities, but not in those without concomitant diseases.

When individual intensive treatments were compared with tooth scaling, scaling and root planing and periodontal flap surgery showed no significant difference relative to tooth scaling (HR: 0.62 [0.31–1.25]) [25]. In contrast, tooth extraction was associated with an increased risk of ischemic stroke (HR: 1.56 [1.19–2.03]) [25].

3.4.5. Combined Dental Treatments

One study reported that the combination of tooth scaling with scaling and root planing or periodontal flap surgery was associated with a lower risk of ischemic stroke compared with tooth scaling alone (HR: 0.82 [0.69–0.98]) [25]. Other combinations of periodontal treatments did not show statistically significant differences compared with tooth scaling monotherapy [25].

3.5. Quantitative Synthesis

After the descriptive analysis, quantitative analyses were performed to estimate the effect of each periodontal health intervention using studies that were sufficiently comparable to be pooled [24, 25, 26, 28, 29, 30]. Due to the heterogeneity across studies, all meta‐analyses were conducted using a random‐effects model.

3.5.1. Any Periodontal Health Intervention

A total of 1,609,327 participants were included in this analysis, with 818,685 in the intervention group and 790,642 in the control group. Periodontal health interventions were associated with a lower risk of ischemic stroke in the general population (HR: 0.88 [0.81–0.95]) (Figure 2). No significant association was observed among individuals over 65 years of age (HR: 0.91 [0.72–1.14]) (Figure S1).

FIGURE 2.

Forest plot for the association between any periodontal health intervention and ischemic stroke (random effects model). The sizes of the grey boxes in the figure are in proportion to the weight assigned to each study, whereas the point within each box is the HR reported by each study. The horizontal bar is the reported 95% CI for each HR.

3.5.2. Tooth Scaling

Tooth scaling was associated with a reduced risk of ischemic stroke (HR: 0.84 [0.76–0.92]) (Figure 3). This association was not statistically significant when analyses were restricted to individuals over 65 years of age (HR: 0.83 [0.57–1.22]) (Figure S2).

FIGURE 3.

Forest plot for the association between tooth scaling and ischemic stroke (random effects model). The sizes of the grey boxes in the figure are in proportion to the weight assigned to each study, whereas the point within each box is the HR reported by each study. The horizontal bar is the reported 95% CI for each HR.

3.5.3. Intensive Periodontal Treatment

Intensive periodontal treatment was not associated with a significant reduction in ischemic stroke risk, either in the general population (HR: 0.96 [0.78–1.18]) (Figure S3) or in special populations (HR: 0.88 [0.69–1.12]) (Figure S4).

3.6. Publication Bias

Visual inspection of funnel plots suggested asymmetry (Figure S5). However, Egger's test showed a low probability of publication bias for all analyses (Table S1).

3.7. Certainty of Evidence

Certainty of the evidence was classified as ‘Very low’ for most of the interventions and ‘Moderate’ for tooth scaling and any periodontal health intervention (Table S2). In these cases, no imprecision or inconsistency was identified, and a strong association was observed. For these interventions, FSN of Rosenberg revealed that the evidence is highly unlikely to change as studies are further added in the future (Effect Size: −0.1440, p < 0.0001, Fail‐safe N: 405 and Effect Size: −0.2014, p < 0.0001, Fail‐safe N: 169, respectively).

4. Discussion

This systematic review and meta‐analysis suggests that periodontal health interventions may play a role in the primary prevention [36] of ischemic stroke in the general population.

An exception to the overall favourable pattern was observed when tooth extraction was considered as part of intensive periodontal treatment [25]. Tooth extraction has been previously proposed as a potential risk factor for cerebrovascular events, possibly due to bacterial translocation and transient systemic inflammation. In contrast, in studies where tooth extraction was combined with other intensive therapies, such as scaling and root planing or periodontal flap surgery, its potential negative effect may have been attenuated [28]. Consequently, the net effect of intensive treatment may appear neutral or even beneficial when heterogeneous procedures are analysed together. These findings emphasize the importance of evaluating intensive periodontal therapies separately in order to distinguish potentially protective interventions from those that may increase cerebrovascular risk.

However, this association is not uniform across all periodontal health interventions. Differences in the type and intensity of the interventions evaluated appear to be a major source of variability among studies, highlighting the importance of examining individual treatments separately rather than considering periodontal care as a homogeneous exposure.

Basic and non‐invasive oral health behaviours, such as regular toothbrushing [24] and routine dental care [24, 27], were associated with a lower risk of ischemic stroke in the included studies. Toothbrushing was assessed according to daily frequency categories, whereas dental visits were defined heterogeneously, ranging from any dental attendance within a given year to more narrowly defined preventive visits, and included consultations with dentists and/or dental hygienists without consistent reporting of the clinical content of the visit. Consequently, both exposures likely capture broader patterns of oral health awareness and preventive health behaviour rather than standardized interventions. Despite the limited number of studies and variability in exposure definitions, the consistency of the observed associations suggests that these oral behaviours may facilitate the prevention or early management of periodontal disease, potentially reducing chronic systemic inflammation and its downstream cerebrovascular effects. Given their low cost, safety and widespread recommendation, these findings strengthen the rationale for promoting good oral hygiene behaviours as a public health strategy.

Tooth scaling showed a statistically significant protective effect against ischemic stroke in the meta‐analysis. However, individual study results differed depending on population characteristics. Studies conducted in patients with diagnosed periodontitis reported a clear benefit [25, 28], whereas those including individuals with unknown periodontal status yielded more heterogeneous results [24, 29]. This suggests that pooling different populations may partially mask population‐specific effects. Nevertheless, the trend of the individual studies suggests that both a meta‐analysis of subpopulations and the selective suppression of each study would achieve similar results, corroborating the protective effect observed in the overall meta‐analysis.

Age also appears to modify the association. The protective effect of tooth scaling was not observed in individuals over 65 years of age in the pooled analysis, although individual studies suggested benefits in specific subgroups, such as older adults over 65 without documented periodontal disease and younger patients with periodontitis.

Overall, tooth scaling seems beneficial both as a preventive measure in periodontally healthy individuals and as a therapeutic intervention in patients with established disease. Nevertheless, further research stratified by age is needed to clarify which subgroups benefit the most, and in the meantime, the limited evidence from individual studies should be taken into account.

In contrast to less invasive interventions, intensive periodontal treatment did not show a significant protective effect against ischemic stroke in the overall quantitative synthesis. One possible explanation is the heterogeneity of procedures included under this category. Tooth extraction, when analysed separately, was associated with an increased risk of ischemic stroke, whereas scaling and root planing or periodontal flap surgery appeared to have neutral or protective effects. Moreover, when comparing each of these procedures with an effective treatment such as tooth scaling, tooth extraction emerges as a risk factor, while scaling and root planing and periodontal flap surgery show protective effects, effectively equalizing the benefit observed with tooth scaling [25]. Likewise, if studies including only tooth extraction were excluded, the overall meta‐analysis would favour intensive interventions, suggesting that the combination of treatments with potentially opposing effects may therefore dilute the overall estimate. These findings highlight the importance of analysing periodontal procedures individually rather than grouping them into broad categories.

Results in special populations were inconsistent. Intensive periodontal treatment was associated with a reduced risk of ischemic stroke in haemodialysis patients [26] but not in individuals with a previous myocardial infarction [30]. Differences in baseline cardiovascular risk, periodontal disease severity, diagnostic accuracy, timing of treatment and post‐event behavioural changes may account for these discrepancies. Additionally, residual confounding due to unmeasured oral hygiene practices or concurrent dental care cannot be excluded. Given these uncertainties, the current evidence does not support a strong recommendation for intensive periodontal treatment solely for stroke prevention in high‐risk populations. Treatment decisions should instead be individualized, prioritizing periodontal severity and overall clinical context.

Limited evidence suggests that combining tooth scaling with more intensive periodontal therapies may offer greater protection against ischemic stroke than tooth scaling alone in patients with established periodontitis [25]. However, this conclusion is based on a single study and should be interpreted with caution, and any application of these findings should be individualized based on patient characteristics. Further well‐designed studies are required to determine whether combined approaches provide incremental benefit over standard care.

The main strengths of this review include the large pooled sample size of 1,886,947 patients, the low risk of bias of the included studies, and the use of quantitative synthesis to support qualitative findings. Nevertheless, several limitations must be acknowledged. Although broad search terms were used to maximize sensitivity, the strategy did not incorporate an extensive range of synonyms or controlled vocabulary and therefore may not be considered fully comprehensive. All included studies were observational, which limits causal inference. There was substantial heterogeneity across studies in terms of intervention type and complexity, patient populations, follow‐up durations and definitions of periodontal disease. This variability was anticipated, given the broad scope of the review, which aimed to assess the association of any periodontal or oral hygiene intervention versus no intervention with ischemic stroke risk. While the heterogeneity limits direct comparability, it was explicitly addressed through random‐effects meta‐analyses and subgroup analyses by intervention type, allowing for the identification of overall trends across diverse interventions and populations. Another limitation is the potential for slight participant overlap in two studies using the 2000 Taiwanese cohort. Although the overlap is expected to be minimal, it should be considered when interpreting the findings. In addition, most studies were conducted in Asian populations, particularly in Taiwan, which may limit the generalizability of the findings to other settings.

5. Conclusions

Results from the present meta‐analysis suggest that periodontal health interventions, particularly tooth scaling, are associated with a reduced risk of ischemic stroke, although differences in how treatments were defined and reported across studies should be taken into account when interpreting the findings.

6. Clinical Relevance

6.1. Scientific Rationale

Several studies have shown an association between periodontitis and ischemic stroke. Based on this evidence, the effects of periodontal health interventions on the incidence of ischemic stroke have been investigated.

6.2. Findings

There is moderate certainty of a favourable effect of periodontal health interventions in general, and tooth scaling specifically, in reducing the risk of ischemic stroke.

6.3. Practical Implications

Those results provide a more solid basis for the formulation of therapeutic recommendations and open avenues for further research into population subgroups, clinical stages, specific treatments and the utility of periodontal disease screening systems.

Author Contributions

Jorge Moldes: contributed to conception, design, data acquisition and interpretation, drafted and critically revised the manuscript. Joao Botelho: contributed to conception, design, data acquisition and interpretation, performed all statistical analyses and critically revised the manuscript. Manuel Rodríguez‐Yáñez: contributed to conception, design and critically revised the manuscript. Vanessa Machado: contributed to conception, design, performed all statistical analyses and critically revised the manuscript. Juan Blanco: contributed to conception, design and critically revised the manuscript. Tomás Sobrino: contributed to conception, design and critically revised the manuscript. Rogelio Leira: contributed to conception, design and critically revised the manuscript. Yago Leira: contributed to conception, design, data acquisition and interpretation, drafted and critically revised the manuscript.

Funding

This study was supported by grants from the Xunta de Galicia (IN607D2024/003). Yago Leira is a recipient of a research contract from the Miguel Servet Program supported by the Institute of Health Carlos III (CP23/00071). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Ethics Statement

The authors have nothing to report.

Consent

The authors have nothing to report.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Any oral health intervention in > 65 years.

Figure S2: Tooth scaling in > 65 years.

Figure S3: Intensive treatment.

Figure S4: Intensive treatment in special populations.

Figure S5: Funnel plots.

Table S1: Results from Egger's test for each periodontal health intervention.

Table S2: GRADE Certainty of Evidence for periodontal health interventions compared to no treatment for the risk of ischemic stroke.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.