Reduced Posterior Occlusal Contacts Are Associated With an Increased Risk of Stroke: A Retrospective Study Based on a Japanese Claims Database

Abstract

Background

Stroke remains a leading cause of mortality and disability worldwide, requiring modifiable risk factors for prevention. Oral health, particularly posterior occlusal contact that supports mastication, may influence systemic vascular outcomes; however, its role in stroke remains unclear. This study investigated whether reduced posterior occlusal contact independently predicts stroke in Japanese adults.

Methods

We conducted a retrospective cohort study using data collected from April 2016 to March 2022 from a nationwide health insurance database in Japan. Adults aged 40 to 74 years without a history of stroke were included. Posterior occlusal status was categorized as Eichner A (full posterior occlusal contact), B (partial contact), or C (no contact). The primary outcome was stroke incidence, identified using validated insurance claims. Cox proportional hazards models were used to estimate hazard ratios (HRs), adjusting for demographics, health behaviors, comorbidities, and tooth counts.

Results

Among 981 543 participants (mean age 49.6±7.0 years; 57.5% male) followed for 2 712 815 person‐years, 7086 strokes occurred. In adjusted Cox models, Eichner B was significantly associated with higher stroke risk than Eichner A (HR, 1.26 [95% CI, 1.10–1.44], P<0.001). Eichner C showed elevated risk, although not statistically significant (HR, 1.17 [95% CI, 0.91–1.50], P=0.229). Associations were observed in both sexes and across age groups.

Conclusions

Reduced posterior occlusal support was associated with higher stroke risk independent of tooth count and conventional vascular risk. These findings highlight that posterior occlusal support may represent a potential marker of cerebrovascular risk and warrant further investigation into its clinical relevance.

Nonstandard Abbreviations and Acronyms

JMDC

Japan Medical Data Center

Clinical Perspective.

What Is New?

• Reduced posterior occlusal contact, a marker of impaired mastication, independently increases stroke risk even after adjustment for vascular risk factors and tooth count.

• The association was most evident in individuals with partial posterior tooth loss, indicating that moderate occlusal decline may have systemic cerebrovascular effects.

What Are the Clinical Implications?

• Posterior occlusal status may represent a modifiable marker for stroke prevention, underscoring the importance of maintaining or restoring occlusal contacts through timely dental care.

Stroke, defined as a sudden neurological deficit caused by acute injury to the central nervous system, remains a leading cause of mortality and long‐term disability worldwide. ¹ , ² Although the age‐standardized incidence and mortality rates have declined in recent decades, the absolute burden of stroke continues to increase, driven by population aging and demographic shifts. ¹ , ² , ³ Ischemic stroke accounts for most cases, followed by intracerebral and subarachnoid hemorrhages. ¹ , ⁴ Although established risk factors such as hypertension, elevated body mass index (BMI), hyperglycemia, and smoking have been extensively studied and targeted in prevention efforts, ¹ , ⁵ other potentially modifiable contributors remain underexplored. Therefore, broadening the scope of stroke prevention requires attention in these domains.

Oral health emerged as a candidate domain with systemic relevance. Several cohort studies have linked poor oral hygiene and tooth loss to an increased stroke risk. ⁶ , ⁷ However, tooth count alone does not fully capture oral function. ⁸ Masticatory ability, which depends not only on the number of teeth but also on their distribution and occlusion, may provide a more meaningful measure of functional oral health. ⁹ Particularly, posterior occlusal contacts are essential for efficient mastication, enabling food comminution, nutrient absorption, and maintenance of systemic health. ¹⁰ , ¹¹ , ¹² Reduced posterior occlusion may be associated with dietary insufficiency, cerebral hypoperfusion, and cognitive decline. ¹³ , ¹⁴ , ¹⁵ , ¹⁶ These findings suggest a plausible pathway linking impaired mastication to an increased cerebrovascular risk. However, no longitudinal studies have directly evaluated whether a reduced posterior occlusal contact can independently predict stroke.

We conducted a large retrospective cohort study using data from a nationwide Japanese claims database. Our objective was to determine whether reduced posterior occlusal contact is associated with an increased stroke risk among adults aged 40 to 74 years, independent of conventional vascular risk factors and tooth count.

METHODS

Data Availability Statement

All data in the Japan Medical Data Center (JMDC) Claims Database are fully anonymized before release; therefore, informed consent from individual participants was not required. The database is available to institutions or researchers who obtain a license from JMDC Inc. (Tokyo, Japan; https://www.jmdc.co.jp/en/), a private company that aggregates nationwide health insurance claims and health checkup data.

Data supporting the findings of this study are available from the corresponding author upon reasonable request. These data are not publicly available due to privacy and ethical restrictions.

Data Source and Study Design

We conducted a retrospective cohort study using the JMDC Claims Database, which includes the medical and dental records of >60 health insurers in Japan. The database covers ∼7.6% of the population, mainly working‐age adults and their dependents under the age of 75 years. It contains encrypted personal IDs, demographics (age and sex), diagnoses, procedures, prescriptions, and annual health checkup results.

We analyzed data from April 2016 to March 2022 to study the association between posterior occlusal contact and stroke (Figure 1). From the 6 549 778 individuals who underwent health checkups, we excluded those aged <40 years (n=2 140 312), those without dental codes for all 28 teeth in the previous year (n=3 418 231), and those with stroke within 12 months before baseline (n=9692). Missing health data were imputed using the mice package in R with 10 data sets. The final cohort included 981 543 individuals with complete dental and health data and no history of stroke. The study design is illustrated in Figure 2.

Figure 1. Sample selection flow chart.

This diagram depicts the stepwise process used to identify eligible participants from the Japan Medical Data Center Claims Database. The initial data set included 6 549 778 individuals with health‐checkup information. The exclusion criteria were as follows: participants younger than 40 years (n=2 140 312), those lacking dental records for all 28 teeth within 1 year before the health check‐up (n=3 418 231), and individuals diagnosed with cardiovascular disease within 12 months before the index date (n=9692).

Figure 2. Overview of the cohort study design.

*Stroke was defined as cerebral infarction, intracerebral hemorrhage, or subarachnoid hemorrhage. ^†Earliest onset of stroke, loss of health insurance coverage, or end of follow‐up.

Outcome

Stroke included cerebral infarction (International Classification of Diseases, Tenth Revision [ICD‐10]: I63), intracerebral hemorrhage (I61), and subarachnoid hemorrhage (I60). ¹⁷ The primary outcome was the first stroke during the follow‐up period (April 2016 to March 2022). We identified cases with confirmed diagnoses from claims and excluded suspected cases. The follow‐up ended at the stroke, death, or study end, whichever occurred first.

Exposure

We used the Eichner classification to assess the posterior occlusal contact. The JMDC Claims Database lacks tooth counts but records dental procedures, such as periodontal examinations and cleanings. Each tooth had a code, and when multiple codes existed, the latest code was used. Following the previous methods, ¹⁸ , ¹⁹ we estimated the number of teeth present in the codes. We identified occlusal support zones, opposing premolars, or molars in contact on each side and grouped them into 0 to 4 zones. Eichner A included all 4 intact zones, Eichner B had 1 to 3 intact zones, and Eichner C had none.

Covariates

We adjusted for demographics, health behaviors, and comorbidities associated with stroke as available in the claims data. The demographics included insurance status (employee or dependent), age, and sex. Age was calculated from the birth date and grouped into 40 to 59 and 60 to 74 years. Health behaviors included obesity, smoking, alcohol use, and physical activity. Obesity was defined as BMI ≥25 kg/m². Smoking and alcohol use were self‐reported; current smokers had smoked >100 cigarettes within the past 28 days, and current drinkers consumed alcohol daily or occasionally. Physical activity was assessed by questionnaire; ideal activity was defined as ≥30 minutes of exercise twice a week or ≥1 hour of walking daily. Tooth count, excluding third molars, was classified as <20 or ≥20 based on evidence that ≥20 teeth are sufficient for functional mastication. ²⁰ Comorbidities included hypertension, hyperglycemia, and dyslipidemia. Hypertension was defined as systolic blood pressure (BP) ≥140 mm Hg, diastolic blood pressure (BP) ≥90 mm Hg, or use of antihypertensive medications. Hyperglycemia was defined as fasting blood glucose level ≥126 mg/dL or use of glucose‐lowering drugs. Dyslipidemia was defined as low‐density lipoprotein ≥140 mg/dL, high‐density lipoprotein <40 mg/dL, triglycerides ≥150 mg/dL, or use of lipid‐lowering drugs. Definitions are listed in Table S1.

Statistical Analysis

All analyses were conducted using R software (version 4.0.4). The statistical significance level was set at α=0.05.

The demographic and clinical characteristics were summarized using descriptive statistics. Continuous variables were expressed as means ± SD and categorical variables as frequencies and percentages. Categorical variables across the 3 Eichner classifications were compared using the chi‐square test. The Anderson–Darling test confirmed that all continuous variables (age, number of teeth, BMI, systolic BP, diastolic BP, fasting blood glucose, low‐density lipoprotein, high‐density lipoprotein, and triglycerides levels) were nonnormally distributed (P<0.001); therefore, we used the Mann–Whitney U test for group comparisons.

Kaplan–Meier curves were plotted to compare stroke incidence among the Eichner groups, and the log‐rank test was used to assess the differences. The proportional hazards assumption was initially evaluated using log–log plots and formally tested using Schoenfeld residuals for each covariate and for the overall model. We used a multivariable Cox proportional hazards model to assess the association between Eichner classification and stroke risk. Hazard ratios (HRs) and 95% CIs were calculated for both unadjusted and adjusted models. Covariates entered simultaneously into the model included age, tooth group, sex, insurance status, smoking, alcohol use, obesity, hypertension, hyperglycemia, dyslipidemia, and physical activity. Martingale residuals were examined to assess the functional form of covariates and to screen for influential observations. Deviance residuals were evaluated to identify outliers and poorly fitted observations. Variance inflation factors were used to assess multicollinearity. For categorical variables, we calculated the adjusted generalized variance inflation factors to align with the standard thresholds. All values were <2.0, indicating no significant multicollinearity.

Four sensitivity analyses were conducted. First, we analyzed only complete cases (n=766 400) and excluded participants with missing health checkup data. Second, we used continuous values (number of teeth, BMI, systolic BP, fasting blood glucose, and low‐density lipoprotein levels) instead of their respective categorical variables. Third, we applied 6‐ and 12‐month induction periods, excluding individuals who developed stroke within these intervals, resulting in sample sizes of 977 688 and 974 792, respectively. Fourth, we defined stroke onset using the ICD‐10 codes I60 to I64 and G459 (n=980 413). Subgroup analyses were stratified by sex (male: n=564 503; female: n=417 040) and age (40–59 years, n=880 299; 60–74 years, n=101 244) to examine the association between Eichner classification and stroke risk.

Ethics

This study used anonymized data from the JMDC Claims Database, in accordance with Japan's Act on the Protection of Personal Information. Informed consent was not required because the data were anonymized under the Next‐Generation Medical Infrastructure Act. According to Japan's ethical guidelines, studies using anonymized data do not require ethics committee approval. The study complied with the principles of the Declaration of Helsinki.

RESULTS

We analyzed 981 543 participants (mean age: 49.6±7.0 years; 564 503 male and 417 040 female) over a total of 2 712 815 person‐years. During a mean follow‐up of 19.5 months, 7086 individuals (mean age: 53.3±7.7 years; 4620 male and 2466 female) developed stroke. Table 1 shows the baseline characteristics according to Eichner classification: A (n=942 444), B (n=28 510), and C (n=10 589). Stroke incidence was significantly higher in Eichner B and C than in Eichner A (P<0.001).

Table 1.

Participants' Characteristics According to the Eichner Classification

	Eichner A (n=942 444)	Eichner B (n=28 510)	Eichner C (n=10 589)	P value*
Stroke, n (%)	6510 (0.7)	403 (1.4)	173 (1.6)	<0.001†
Age, y	49.2 (6.7)	58.3 (7.8)	61.8 (6.9)	<0.001‡
60–75 y, n (%)	81 075 (8.6)	13 207 (46.3)	6962 (65.7)	<0.001†
Male sex, n (%)	538 196 (57.1)	18 828 (66.0)	7439 (70.3)	<0.001†
Employee, n (%)	779 971 (82.8)	24 165 (84.8)	9363 (88.4)	<0.001†
Body mass index	23.1 (3.8)	23.8 (3.8)	23.6 (3.8)	<0.001‡
Obesity, n (%)	248 838 (26.4)	9560 (33.5)	3357 (31.7)	<0.001†
Hypertension, n (%)	203 767 (21.6)	12 044 (42.2)	5307 (50.1)	<0.001†
Hyperglycemia, n (%)	43 712 (4.6)	3784 (13.3)	2017 (19.0)	<0.001†
Dyslipidemia, n (%)	354 901 (37.7)	13 463 (47.2)	5132 (48.5)	<0.001†
Current smoking, n (%)	183 870 (19.5)	9371 (32.9)	4684 (44.2)	<0.001†
Alcohol, n (%)	234 526 (24.9)	9366 (32.9)	3745 (35.4)	<0.001†
Physical inactivity, n (%)	582 879 (61.8)	16 944 (59.4)	6404 (60.5)	<0.001†
Systolic blood pressure, mm Hg	119.6 (16.0)	126.7 (17.1)	129.6 (18.0)	<0.001‡
Diastolic blood pressure, mm Hg	74.9 (11.9)	77.9 (11.4)	78.2 (11.6)	<0.001‡
Low‐density lipoprotein, mg/dL	124.0 (30.7)	124.4 (31.5)	123.1 (32.8)	<0.001‡
High‐density lipoprotein, mg/dL	65.3 (17.4)	63.5 (17.8)	62.2 (18.0)	<0.001‡
Triglycerides, mg/dL	106.3 (84.2)	121.4 (94.2)	129.3 (100.0)	<0.001‡
Fasting glucose level, mg/dL	95.5 (16.2)	102.4 (23.2)	106.3 (27.0)	<0.001‡
Number of teeth	27.9 (0.3)	20.0 (6.0)	4.1 (3.3)	<0.001‡
1–19 teeth, n (%)	82 (0.0)	10 974 (38.5)	10 589 (100.0)	<0.001†

Occlusal conditions were categorized according to the Eichner classification, which is determined by the number of posterior support zones (0–4). A: contacts present in all 4 posterior support zones; B: contacts limited to 1 to 3 posterior support zones or restricted to the anterior teeth; C: no posterior contacts identified.

^*Differences among the Eichner A, B, and C groups.

^† P values for chi‐square test.

^‡ P values for ANOVA.

Diagnostic analyses revealed no meaningful violations of the assumptions of the Cox proportional hazards model. The global Schoenfeld test was not significant (P=0.55), and no consistent time‐dependent patterns were observed (Table S2). Although smoking showed a borderline covariate‐specific result, additional assessment of its Schoenfeld residuals indicated no meaningful time dependence (Figure S1). Martingale residuals showed the expected distribution without evidence of nonlinearity (Figure S2), and deviance residuals identified few extreme observations (0.3%) with no indication of undue influence (Figure S3).

Figure 3 presents the Kaplan–Meier curves of stroke incidence according to the Eichner classification. Stroke risk showed significant difference among the 3 groups (log‐rank P<0.001). In the Cox models adjusted for demographics, health behaviors, comorbidities, and tooth count, the Eichner B score was significantly associated with stroke risk (Table 2 and Table S3). Eichner B had a 1.26‐fold higher risk than Eichner A (HR, 1.26 [95% CI, 1.10–1.44], P<0.001), whereas Eichner C showed a 1.17‐fold increase (HR, 1.17 [95% CI, 0.91–1.50], P=0.229).

Figure 3. Kaplan–Meier curves for stroke‐free survival according to Eichner classification (n=981 543).

Table 2.

Univariable and Multivariable Cox Proportional Hazards Model for Stroke (n=981 543)

	Univariable			Multivariable
	HR	95% CI	P value	HR	95% CI	P value
Eichner classification
Eichner A	1 (reference)			1 (reference)
Eichner B	2.27	2.06–2.52	<0.001	1.26	1.10–1.44	<0.001
Eichner C	2.73	2.35–3.18	<0.001	1.17	0.91–1.50	0.229

Occlusal conditions were classified into 3 groups: A (posterior contacts in all four support zones), B (posterior contacts limited to 1 to 3 support zones or limited to the anterior teeth), and C (no posterior contacts). Adjusted HRs and 95% CIs were estimated using Cox proportional hazards models, with all covariates entered simultaneously, including tooth group, age, sex, insurance status, smoking, alcohol use, obesity, hypertension, hyperglycemia, dyslipidemia, and physical activity. HR indicates hazard ratio.

Several sensitivity analyses were conducted to assess the robustness of the primary findings. First, an analysis limited to complete cases without multiple imputations produced results consistent with the main analysis (Table S4). Second, replacing the categorical variables with continuous values for tooth count, BMI, hypertension, hyperglycemia, and dyslipidemia yielded similar results (Table S5). Third, we applied 6‐ and 12‐month induction periods to exclude participants with potential latent stroke at baseline, and the associations remained stable (Table S6). Finally, the redefinition of stroke using alternative ICD‐10 codes did not alter our findings (Table S7). Together, these analyses confirmed the robustness of the association between posterior occlusal status and stroke across different model specifications and populations.

We also examined the association between the Eichner classification and stroke risk according to sex and age. In men, the Eichner B had a significantly higher risk of stroke than the Eichner A (HR, 1.24 [95% CI, 1.05–1.45], P=0.010), and a similar association was observed in women (HR, 1.31 [95% CI, 1.03–1.67], P=0.027). Age‐stratified analyses showed that in the 40 to 59 age group, the Eichner B group had a significantly increased risk (HR, 1.27 [95% CI, 1.04–1.55], P=0.021). In the 60–74 age group, both Eichner B (HR, 1.31 [95% CI, 1.09–1.57], P=0.003) and C (HR, 1.39 [95% CI, 1.02–1.90], P=0.038) were significantly associated with higher stroke risk than Eichner A (Tables S8 and S9).

DISCUSSION

In this longitudinal study, we used a nationwide Japanese claims database to examine the association between posterior occlusal contact, measured using the Eichner classification, and stroke risk. After adjusting for conventional vascular risk factors and tooth count, reduced posterior occlusal contact was associated with a higher risk of incident stroke compared with full occlusion. This association was most consistent with the partial loss of posterior support (Eichner B). Complete loss of posterior contact (Eichner C) showed a directionally similar but nonsignificant overall increase. Importantly, age‐stratified analyses further indicated that these associations were more prominent among older adults, particularly those aged 60 to 74 years, where baseline stroke incidence was higher and event rates allowed more stable effect estimation. The sensitivity analyses supported the robustness of these findings.

To our knowledge, this is the first longitudinal study to show that reduced posterior occlusal contact, a key determinant of mastication, independently increases the risk of stroke. Prior research has not identified impaired chewing as a risk factor for stroke, but has instead shown that stroke often impairs chewing, oral sensitivity, and mastication due to neurophysiological damage in the orofacial region. ²¹ , ²² , ²³ , ²⁴ , ²⁵ For example, one cohort study found no improvement in chewing efficiency after poststroke dental restoration, and no correlation between occlusal units and mastication in stroke survivors. ²⁴ These findings suggest that poststroke masticatory decline stems mainly from neurological causes rather than dental issues. ²⁶ , ²⁷ In contrast, many studies have associated tooth loss with a higher risk of stroke. ⁷ , ²⁸ , ²⁹ Because mastication depends on both the number and distribution of occluding teeth, particularly in the posterior region, posterior tooth loss can impair functional occlusion and chewing efficiency. ³⁰ Notably, stroke risk remained significantly elevated in individuals classified as Eichner B, even after adjusting for the number of remaining teeth. This suggests that the partial loss of posterior occlusal support, rather than tooth loss alone, plays a distinct role in increasing stroke risk. A significant interaction was observed between Eichner B and the number of remaining teeth, indicating that the impact of tooth loss is modified by the degree of occlusal support (Table S10). The association in Eichner C, which reflects a complete loss of posterior occlusion, weakened and lost significance after adjusting for tooth count. One possible explanation for this is the varying effects of denture use. Many patients in the Eichner C group were likely to use dentures; however, their health effects remain uncertain. Removable partial dentures can restore mastication and improve quality of life, ³¹ however, denture use increases the risk of cardiovascular diseases, including stroke. ³² As the present data set lacked information on denture status, we could not resolve these conflicting findings. This limitation underscores the need for future studies including data on prosthetic treatments.

The association between reduced posterior occlusal support and stroke risk is biologically and clinically plausible, particularly considering the established role of mastication in neurovascular regulation. Chewing activates trigeminal afferents and stimulates cortical regions, such as the prefrontal cortex and hippocampus, thereby enhancing cerebral perfusion and oxygenation. ³³ , ³⁴ , ³⁵ Neuroimaging and experimental studies have shown that diminished mastication, which often results from tooth loss or posterior occlusal decline, is associated with cognitive impairment and impaired neurovascular coupling. ³⁶ , ³⁷ Consistent with these findings, our previous study showed that individuals with fewer posterior occlusal contacts as defined by the Eichner classification were at a higher risk of dementia. ¹⁶ Together, these results support the notion that the loss of sensory input from mastication may compromise cerebral regulation. As cognitive decline and stroke share common pathophysiological mechanisms, including endothelial dysfunction, blood–brain barrier disruption, and impaired cerebral autoregulation, ³⁸ reduced masticatory stimulation could possibly increase stroke risk through the neurovascular and endothelial pathways. In addition to the neurovascular mechanisms, dietary alterations secondary to impaired mastication represent another plausible pathway linking posterior occlusal loss to stroke. Individuals with reduced occlusion often avoid hard‐to‐chew foods, such as fibrous vegetables, fruits, and lean proteins, leading to diets that are less diverse and nutritionally unbalanced. ³⁹ , ⁴⁰ , ⁴¹ Such dietary changes decrease the intake of cardioprotective nutrients, including antioxidants, fiber, and omega‐3 fatty acids, which play critical roles in the prevention of atherosclerosis and hypertension. ⁴⁰ , ⁴² Consequently, compromised masticatory ability may indirectly contribute to the risk of stroke by promoting adverse dietary patterns that deteriorate vascular health.

Reduced posterior occlusal contacts were significantly associated with cerebral infarction but not with intracerebral or subarachnoid hemorrhage (Table S11). This pattern is consistent with previous studies showing that tooth loss and dental caries increase the risk of cerebral infarction but not hemorrhagic stroke. ⁶ These differences suggest that oral health has a weaker influence on the mechanisms underlying hemorrhagic stroke. Cerebral infarction is closely linked to atherosclerosis and chronic inflammation, and is modulated by poor oral hygiene and periodontal disease. ⁴³ , ⁴⁴ In contrast, intracerebral and subarachnoid hemorrhages most commonly result from the sudden rupture of blood vessels, typically due to chronic hypertension or congenital vascular anomalies, such as aneurysms and arteriovenous malformations. ⁴⁵ , ⁴⁶ Therefore, the inflammatory and atherogenic pathways that plausibly connect impaired occlusion to infarction are unlikely to contribute substantially to hemorrhagic stroke, which explains the lack of an association in our findings.

This study has several limitations that should be considered when interpreting the findings. First, the diagnostic data were based on insurance claims that were not routinely validated against clinical records. Misclassification of stroke diagnoses may have occurred, which could have affected the accuracy of the outcome assessment. Second, this retrospective cohort study identified associations but could not establish causality. Although we found a significant link between posterior occlusal status and stroke, the observational design requires cautious interpretation. Prospective studies with standardized assessments are required to confirm these findings. Third, the study population was limited to working‐age adults from large Japanese companies and their dependents. Older adults (>75 years) were underrepresented, further limiting the generalizability of the study. This age structure also affects the interpretation of the overall estimates, because stroke incidence was low among younger participants and the relatively younger, lower‐risk composition of the reference group may have influenced the magnitude of the effect estimates despite statistical adjustment. The data set may also lack sufficient geographic and socioeconomic diversity, which could introduce a selection bias. Fourth, although we adjusted for common stroke risk factors, unmeasured confounders, such as income, education, occupation, diet, and health care access, may have affected the results. Finally, we lacked data on periodontal status despite its known relation to inflammation and stroke risk. ⁴⁷

Conclusions

In conclusion, this observational study demonstrated that a reduced posterior occlusal contact, as defined by the Eichner classification, was associated with an increased risk of stroke, with the association more evident in older adults. Because of the observational design and relatively short median follow‐up, these findings should be considered hypothesis generating rather than causal. Nevertheless, the findings underscore the potential significance of posterior occlusal support as a plausible factor relevant to cerebrovascular risk in aging populations, warranting further examination in prospective and interventional studies to clarify underlying mechanisms and determine whether improving oral rehabilitation influences long‐term cerebrovascular outcomes.

Sources of Funding

This study was funded by Grants‐in‐Aid for Scientific Research (22H03299, 23K24557) from the Japan Society for the Promotion of Science KAKENHI and Health and Labor Sciences Research Grants (24FA1020) from the Ministry of Health, Labor, and Welfare.

Disclosures

None.

Supporting information

Tables S1–S11

Figures S1–S3