Development and Validation of a Risk Prediction Model for Oral Frailty in Older Adults With Mild Cognitive Impairment: A Cross-Sectional Study

Development and Validation of a Risk Prediction Model for Oral Frailty in Older Adults With Mild Cognitive Impairment: A Cross-Sectional Study | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Development and Validation of a Risk Prediction Model for Oral Frailty in Older Adults With Mild Cognitive Impairment: A Cross-Sectional Study Yangkun Zhang, Jingjing Wang, Xiaoming Liu, Jing Li This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8962009/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 5 You are reading this latest preprint version Abstract Objectives Oral frailty is a multidimensional geriatric syndrome associated with adverse health outcomes. Older adults with mild cognitive impairment (MCI) may be particularly vulnerable; however, prediction tools tailored to this population remain limited. This study aimed to develop and internally validate a risk prediction model for oral frailty in older adults with MCI. Design Cross-sectional study for model development. Setting and Participants A total of 456 older adults diagnosed with MCI were recruited from a geriatric outpatient clinic between October 2025 and January 2026. Participants were randomly divided into a training cohort (70%) and an independent validation cohort (30%). Methods Oral frailty was assessed using the Oral Frailty Index-8. Candidate predictors included demographic, clinical, nutritional, and psychosocial variables. Feature selection was performed using random forest and least absolute shrinkage and selection operator regression. Logistic regression, decision tree, and extreme gradient boosting models were developed. Model discrimination, calibration, and clinical utility were evaluated in the validation cohort. Results The prevalence of oral frailty was 73.9%. In the validation cohort, the logistic regression model demonstrated good discrimination (AUC 0.917, 95% CI 0.855–0.979) and satisfactory calibration. Decision curve analysis indicated favorable net clinical benefit across a range of threshold probabilities. Independent predictors included denture burden (OR 161.016, 95% CI 42.603–920.001), chewing difficulty (OR 12.055, 95% CI 4.483–36.770), malnutrition (OR 8.076, 95% CI 2.909–25.738), depressive symptoms (OR 0.665, 95% CI 0.484–0.893), and living arrangement (OR 0.047, 95% CI 0.003–0.545). Conclusions and Implications Oral frailty among older adults with MCI is strongly associated with structural oral impairment, functional decline, nutritional vulnerability, and psychosocial factors. The proposed model demonstrates good predictive performance and may facilitate early identification of high-risk individuals in clinical settings. Integrated oral–cognitive assessment strategies may improve comprehensive geriatric care for this vulnerable population. Oral frailty oral health mild cognitive impairment cognitive impairment older adults Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction With the accelerating process of global population aging, the coexistence of geriatric syndromes has emerged as a major public health challenge. 1 Mild cognitive impairment (MCI), a transitional clinical stage between normal cognitive aging and dementia, is characterized by memory decline and executive dysfunction. 2 Its prevalence increases substantially with advancing age. 3 By 2020, an estimated 38.77 million older adults aged 60 years and above in China were affected by MCI. 4 Beyond its well-established role as a precursor to dementia, MCI is frequently accompanied by systemic functional decline, significantly compromising independence and quality of life. 5, 6 Oral health has increasingly been recognized as a fundamental component of overall health in older adults. 7 Oral frailty is an age-related syndrome involving progressive deterioration in oral structure and function, manifesting as chewing difficulty, swallowing dysfunction, and tooth loss. 8 Oral frailty not only impairs nutritional intake and daily functioning but may also interact bidirectionally with cognitive decline. 9, 10 Older adults with cognitive impairment often experience diminished memory, executive function, and fine motor coordination, which reduce their ability to maintain adequate oral hygiene and increase susceptibility to dental caries, periodontal disease, and tooth loss. 11 Conversely, chronic oral inflammation may serve as a persistent source of systemic low-grade inflammation. 12 Pro-inflammatory cytokines such as tumor necrosis factor-α and interleukin-6 can cross the blood–brain barrier, promote neuroinflammation, and accelerate β-amyloid deposition, thereby exacerbating cognitive deterioration. 13, 14 This bidirectional interaction suggests that the coexistence of MCI and oral frailty may generate synergistic adverse effects, increasing the risk of falls, malnutrition, disability, and mortality, and imposing substantial burdens on families and healthcare systems. 8, 15, 16 Emerging evidence indicates that older adults with MCI are at significantly higher risk of developing oral frailty compared with cognitively intact peers. 17 When these conditions coexist, the risk of adverse health outcomes is markedly amplified, presenting complex challenges for geriatric care. Despite this growing recognition, current prediction models for oral frailty have primarily focused on community-dwelling older populations or older adults with chronic disease, and have not specifically addressed the unique characteristics of individuals with MCI. 18 , 19, 20 Cognitive impairment may obscure symptom reporting, reduce self-care capacity, and alter health behaviors, thereby necessitating targeted risk stratification tools for this vulnerable group. 21 Moreover, clinical management of MCI has traditionally emphasized delaying progression to dementia, whereas systematic assessment and integrated intervention for oral health remain underexplored. 22 Machine learning techniques have demonstrated strong capability in handling multidimensional clinical data and identifying complex nonlinear associations. These methods have achieved promising performance in risk prediction for cardiovascular disease, diabetes, and other chronic conditions. 23 However, studies applying machine learning to predict oral frailty specifically among older adults with MCI remain limited. Therefore, this study aimed to develop and internally validate a risk prediction model for oral frailty in older adults with MCI using multiple machine learning algorithms. By integrating multidimensional clinical indicators, we sought to enable early identification of high-risk individuals and provide a foundation for personalized, proactive oral health interventions. Ultimately, this work may contribute to the development of an integrated cognitive–oral health management framework and improve health outcomes in this vulnerable population. Methods Study Design This was a prospective observational study conducted to develop and internally validate a clinical prediction model for oral frailty among older adults with MCI. Participants Participants were recruited using convenience sampling from the Geriatric Outpatient Department of General Hospital of Northern Theater Command between October 2025 and January 2026. Participants were eligible if they: (1) were aged ≥60 years; (2)met diagnostic criteria for mild cognitive impairment; (3)had stable physical conditions and were able to complete questionnaires and oral examinations; (4)provided written informed consent. Participants were excluded if they: (1)had severe psychiatric disorders impairing cooperation; (2)had oral and maxillofacial deformities or tumors; (3)had end-stage systemic diseases (eg, advanced hepatic or renal failure, malignancy). Sample size was estimated according to the TRIPOD recommendations for prediction model development.Based on the events-per-variable (EPV) principle, at least 10 outcome events per predictor variable were required to ensure model stability. Assuming an estimated oral frailty prevalence of 57% from prior literature and considering up to 24 candidate predictors, the minimum required sample size was calculated as 421 participants. 24 After accounting for an anticipated 5% attrition rate, the final target sample size was set at 442 participants. Ethical approval was obtained from the Ethics Committee of General Hospital of Northern Theater Command [ Approval No: Y (2026) 08 ].Written informed consent was obtained from all participants prior to enrollment in the study. Measurements Sociodemographic and Clinical Data A structured questionnaire developed based on literature review and expert consultation was used to collect demographic and clinical information.Collected variables included:sex, age, body mass index (BMI), marriage, education level,income, medical insurance, residential setting, living arrangement, smoking, drinking, comorbidity,dentures. Assessment of Cognition In the present study, mild cognitive impairment (MCI) was diagnosed according to the criteria established by the Alzheimer’s Association McKhann and Albert 2011 criteria. 25 The diagnostic criteria were as follows: (1) for participants aged 60~79 years, 80~89 years, and ≥90 years, cognitive function was assessed using the Peking Union Medical College Hospital version of the Montreal Cognitive Assessment (MoCA-P). 26 The optimal cutoff values for identifying mild cognitive impairment were ≤25, ≤24, and ≤23, respectively, across these three age groups; (2) possible mild impairment in complex instrumental activities of daily living, while basic activities of daily living remained functionally independent; (3) absence of a diagnosis of dementia; and (4) confirmation of the diagnosis by a senior attending neurologist at a tertiary hospital. Assessment of Oral Frailty Oral frailty among older adults with MCI was assessed using the Oral Frailty Index-8 Scale (OFI-8) developed by Tanaka et al. 27 The scale comprises five domains: denture use, swallowing function, social participation, oral health behaviors, and masticatory function, encompassing a total of eight items. Each item is answered dichotomously (“yes” or “no”). The first three items are scored from 0 to 2 points, whereas the remaining five items are scored from 0 to 1 point, yielding a total score ranging from 0 to 11. A total score ≥4 indicates oral frailty, a score of 3 denotes prefrailty, and a score of 0~2 indicates no oral frailty. Assessment of Sarcopenia Sarcopenia status was evaluated using the Sarcopenia-Five (SARC-F) questionnaire developed by Saint Louis University. 28 The scale consists of five items assessing strength, assistance in walking, rising from a chair, climbing stairs, and falls. Each item is scored from 0 to 2 points, with a total score ranging from 0 to 10. A total score ≥4 indicates a high risk of sarcopenia. Assessment of Depression Depressive symptoms were assessed using the 15-item Geriatric Depression Scale (GDS-15) developed by Yesavage et al. 29 The scale comprises two dimensions and includes 15 dichotomous (“yes” or “no”) items. Total scores range from 0 to 15, with higher scores indicating more severe depressive symptoms. A score ≥5 is considered indicative of clinically significant depressive symptoms. Assessment of Malnutrition Nutritional status was assessed using the Mini Nutritional Assessment–Short Form (MNA-SF) developed by Rubenstein et al. 30 The instrument includes six items evaluating BMI, recent weight loss, medical history, mobility, neuropsychological status, and dietary intake. Total scores range from 0 to 14, with scores ≥11 indicating normal nutritional status and scores <11 indicating risk of malnutrition or malnutrition. Data Collection Procedures Data were collected through face-to-face interviews and clinical examinations. A multidisciplinary research team consisting of geriatricians, dental physicians, geriatric nurses, and public health experts underwent standardized training before study initiation. Eligible participants were screened according to inclusion and exclusion criteria. After obtaining informed consent, trained researchers conducted one-on-one structured interviews and clinical oral examinations. All questionnaires were reviewed on-site to ensure completeness. Oral examinations and sarcopenia assessments were performed according to standardized clinical procedures. Statistical Analysis All statistical analyses were conducted using R software (version 4.5.2, R Foundation for Statistical Computing). A two-sided P value <0.05 was considered statistically significant.Continuous variables were assessed for normality using the Shapiro–Wilk test. As most continuous variables were non-normally distributed, they were expressed as median and interquartile range (IQR) and compared using the Mann–Whitney U test. Categorical variables were presented as frequencies and percentages and compared using the chi-square test or Fisher’s exact test as appropriate. No missing data were observed in the variables included in the analysis. Model Development Strategy The primary objective of this study was to develop and internally validate a prediction model for oral frailty among older adults with MCI. Stratified random sampling based on outcome status was applied to maintain similar event proportions across datasets.The entire cohort was randomly divided into a training dataset (70%) for model development and a validation dataset (30%) for independent performance evaluation. Feature Selection A two-stage feature selection process was implemented to reduce dimensionality and minimize overfitting.First, variables demonstrating statistical significance in univariate analyses were entered into a random forest model to estimate relative variable importance.Second, least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation was performed to identify the most predictive subset of variables. The lambda.1se criterion was adopted to select a parsimonious model while maintaining stability.Variance inflation factors (VIFs) were examined to assess multicollinearity among retained predictors. Model Construction Based on the selected predictors, three classification models were developed:Binary logistic regression; Decision tree; Extreme gradient boosting (XGBoost). Model development was conducted in the training dataset. Model performance was primarily evaluated in the independent validation dataset. Model Performance Evaluation Model discrimination was assessed using the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and accuracy were calculated based on predicted probabilities. 31 The optimal probability threshold was determined using the Youden index. Model calibration was evaluated using:Hosmer–Lemeshow goodness-of-fit test,calibration plots,calibration slope and intercept. 32 These metrics assessed agreement between predicted probabilities and observed outcomes.Overall predictive accuracy was evaluated using the Brier score, which measures the mean squared difference between predicted probabilities and observed outcomes.Internal validation was performed using bootstrap resampling (1,000 repetitions) in the training dataset to estimate optimism-corrected performance measures.The events-per-variable (EPV) ratio was calculated to assess model stability.Decision curve analysis (DCA) was performed to evaluate the net clinical benefit of the final model across a range of threshold probabilities. 32 Sensitivity Analysis To assess the influence of dentures on model performance, a sensitivity analysis was conducted by excluding dentures from the model. The resulting AUC was compared with that of the primary model. Results Participant Characteristics A total of 500 older adults who met the diagnostic criteria for mild cognitive impairment (MCI) were screened. After excluding 28 individuals who declined participation and 16 who withdrew during follow-up, 456 participants were included in the final analysis. Among them, 337 were identified as having oral frailty according to the OFI-8 criteria, corresponding to a prevalence of 73.9%. The median age of participants was 68.0 years (interquartile range [IQR], 63.0–72.0), and 54.8% were male. The cohort was randomly divided into a training set (n = 319, 236 events) and an independent validation set (n = 137, 101 events). No statistically significant differences in baseline characteristics were observed between the 2 datasets (Supplementary Table 1), indicating appropriate distributional balance. Univariate analysis of oral frailty Continuous variables were standardized ; categorical variables were dummy-coded. Between-group comparisons revealed significant differences in age, educational level, residential setting, living arrangement, smoking status, comorbidities, sarcopenia, malnutrition, chewing difficulty, denture, and depressive symptoms (all P < 0.05). Detailed results are presented in Table 1. Variable Selection A two-stage feature selection strategy was applied. First, variables with statistical significance in univariate analyses were entered into a random forest model to estimate relative feature importance. Subsequently, the top-ranked predictors were subjected to least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation. The lambda.1se criterion was adopted to obtain a parsimonious model while minimizing overfitting. Nine predictors were retained for model construction: denture, chewing difficulty, sarcopenia, malnutrition, depressive symptoms, educational level, smoking status, comorbidities, and living arrangement.As shown in Figure 1 and Figure 2. Multicollinearity among retained predictors was assessed using variance inflation factors (VIFs). All VIF values were below 2, indicating no evidence of significant multicollinearity.Pairwise correlations among predictors were low, absolute Spearman correlation coefficients < 0.30 (Supplementary Figure1). The events-per-variable ratio exceeded 30 (337 events / 9 predictors), indicating adequate model stability. Model Comparison Three classification models were developed: binary logistic regression, decision tree, and extreme gradient boosting (XGBoost). Model performance was primarily evaluated in the validation dataset.The comparative performance results are summarized in Table 2.Although the three classification models demonstrated comparable discrimination in the validation dataset, the binary logistic regression model was selected as the final prediction model due to its balanced sensitivity, interpretability, and clinical applicability. While XGBoost showed slightly higher specificity, logistic regression provided superior sensitivity and comparable AUC performance, supporting its use as the primary model. Final Model Performance Discrimination In the validation dataset, the final logistic regression model achieved an AUC of 0.917 (95% CI: 0.855–0.979), indicating excellent discrimination. The corresponding training AUC was 0.965 (95% CI: 0.946–0.984). The ROC curve of the validation set is shown in Figure 3.These findings suggest strong ability to distinguish patients with and without oral frailty. Calibration Calibration was assessed using the Hosmer–Lemeshow goodness-of-fit test and calibration plots. The Hosmer–Lemeshow test indicated no evidence of poor fit ( x ² = 6.44, df = 8, P = 0.598). The calibration slope and intercept suggested acceptable agreement between predicted and observed probabilities. Calibration curves confirmed adequate model calibration(Figure 4). Internal Validation Internal validation using 1,000 bootstrap resamples yielded an optimism-corrected AUC of 0.945 in the training dataset, suggesting minimal overfitting. Overall Accuracy The Brier score was 0.057 in the training set and 0.100 in the validation set, indicating good overall predictive accuracy. Clinical Utility Decision curve analysis demonstrated that the final model provided greater net benefit across a broad range of threshold probabilities compared with treat-all and treat-none strategies, supporting its potential clinical applicability(Figure 5). Optimal Threshold This study determined the optimal threshold using the Youden index rather than the conventional 0.5 cutoff.The optimal probability threshold determined using the Youden index was 0.696, corresponding to a sensitivity of 0.911 and specificity of 0.964 in the validation dataset. Logistic Regression Estimates Multivariable logistic regression identified denture, chewing difficulty, malnutrition, living arrangement, and depressive symptoms as independent predictors of oral frailty (Table 3). Denture demonstrated the strongest association with oral frailty (OR 161.02; 95% CI, 42.60~920.00; P < 0.001). Sensitivity Analysis To assess the contribution of denture number to overall model performance, a sensitivity analysis excluding this variable was conducted. In the validation dataset, the AUC decreased to 0.797(Supplementary Figure 2), indicating that denture burden substantially contributes to the model’s discriminative ability. Discussion In this prospective cohort of older adults with mild cognitive impairment (MCI), we developed and internally validated a multidimensional risk prediction model for oral frailty that demonstrated robust discrimination and calibration performance. The identification of dentures, chewing difficulty, malnutrition, depressive symptoms, and living arrangement as influential predictors reinforces the concept of oral frailty as a systemic geriatric syndrome rather than an isolated oral disorder. Our findings highlight the intertwined structural, functional, nutritional, and psychosocial determinants of oral vulnerability in cognitively impaired populations and provide a clinically pragmatic framework for early risk stratification. In this study, the prevalence of oral frailty among older adults with mild cognitive impairment was 73.9%, underscoring the complex and potentially bidirectional relationship between oral frailty and cognitive impairment in later life. Population-based longitudinal data have shown that oral frailty significantly predicts the risk of new-onset MCI among community-dwelling older individuals, even after adjustment for confounding factors such as physical frailty and demographic characteristics. 15 Beyond incidence associations, the pathophysiological mechanisms connecting oral frailty and cognitive deterioration are increasingly supported by epidemiological and mechanistic studies. Declining oral function may reduce masticatory stimulation and sensory input that support neurocognitive integrity, while chronic oral inflammation can contribute to systemic inflammation and neuroinflammatory cascades that accelerate cognitive impairment. 33 Clinicians should consider early screening using OFI-8 in combination with the MoCA to facilitate timely identification of individuals at elevated risk. Integration of salivary and blood-based biomarkers may further enhance risk stratification and support precision identification of high-risk populations. 8 Interventions should prioritize control of inflammatory burden and functional rehabilitation. Comprehensive periodontal therapy, timely prosthetic rehabilitation for tooth loss, and standardized management of xerostomia are essential to reduce chronic oral infection and inflammation. 34 These measures should be complemented by masticatory function training and nutritional support—particularly adequate protein intake—to mitigate the development of oral frailty and potentially slow the progression of cognitive decline in older adults. In the context of clinical assessment, our selection of denture number as a graded indicator rather than a binary variable reflects the clinical reality that severity of tooth loss and prosthetic burden may more accurately represent oral function decline. 35 The number of dentures better captures the magnitude of oral functional loss and was highly predictive in our model.Although the magnitude of the estimated odds ratio was substantial, this likely reflects the categorical structure of the variable and distribution imbalance between exposure groups rather than a literal 160-fold biological increase in risk. Extensive tooth loss and reliance on multiple dentures may serve as markers of cumulative structural oral deterioration, compromised occlusal stability, and reduced masticatory efficiency. 36 However, the sensitivity analysis excluding denture number demonstrated reduced discrimination (AUC = 0.797), indicating that although denture burden carries strong predictive signal, it should be interpreted in the context of comprehensive oral function rather than as a sole defining criterion. This aligns with emerging views that multidimensional indicators of oral status, such as chewing difficulty, swallowing impairment, and salivary function, may offer more nuanced insights into overall oral frailty and its relationship with systemic outcomes. 37 Notably, several risk factors identified in our model, including chewing difficulty, malnutrition, depressive symptoms, and living arrangement have also been implicated as determinants of both oral frailty and cognitive dysfunction. Mastication represents a critical functional link between oral structures and systemic health. 38 Chewing difficulty may reduce dietary diversity and protein intake, thereby increasing the risk of malnutrition. 39 In turn, malnutrition can compromise oral mucosal integrity, delay wound healing, and weaken the muscular strength required for effective swallowing. 40 Given the limited adherence to oral self-care commonly observed in individuals with MCI, a caregiver-led oral care model should be considered in clinical practice. Standardized chewing rehabilitation programs aimed at improving muscle strength and coordination may help prevent further functional decline. In addition, individualized dietary plans tailored to patients’ graded chewing capacity are recommended to avoid monotonous diets and nutritional deficiencies resulting from chewing limitations. Such comprehensive strategies may help delay the progression of sarcopenia and physical frailty and potentially mitigate downstream contributors to cognitive decline. 41 Depressive symptoms were significantly associated with oral frailty, suggesting that oral frailty should not be regarded solely as a functional oral condition. Rather, it may represent a broader geriatric frailty phenotype that encompasses emotional and psychosocial dimensions. 42 Living arrangement emerged as an independent predictor of oral frailty, suggesting that the social environment may substantially influence oral health behaviors, dietary patterns, and access to care. 43 Older adults with MCI who live with a spouse or adult children may benefit from shared meals, informal health monitoring, and caregiving support, thereby reducing the risk of developing oral frailty. Our study adds to the growing literature advocating for integrated predictive tools in geriatric practice. Whereas most existing oral frailty research has focused on prevalence and cross-sectional associations, there remains a paucity of clinical tools specifically tailored to populations with cognitive impairment. Prior models developed for general older adult populations may not account for the unique behavioral and functional characteristics of individuals with MCI, such as reduced self-care capacity and atypical symptom reporting. Our model addresses this gap by leveraging machine learning-assisted variable selection to capture multidimensional relevance while maintaining interpretability through logistic regression—an approach well-aligned with clinical decision-making needs. Limitations Several limitations merit consideration. First, the single-center design and the lack of external validation may limit generalizability. Future multicenter studies with external cohorts are needed to confirm the transportability of our model. Second, although our model demonstrated promising performance metrics, the OFI-8 instrument primarily relies on self-reported items and screening responses; incorporation of objective oral health measures, such as tongue pressure or periodontal indices, may further enhance predictive utility. Finally, while the prediction model reflects associations rather than causal relationships, it provides clinically actionable risk stratification that may guide targeted oral health and nutritional interventions in older adults with MCI. Conclusion Our study presents a clinically relevant and statistically robust prediction model for oral frailty in older adults with MCI. By integrating clinical, nutritional, and psychosocial indicators, the model facilitates early identification of high-risk individuals, which may inform tailored preventive strategies. Given the growing prevalence of both cognitive impairment and oral frailty in aging populations, such predictive tools hold promise for guiding interdisciplinary interventions aimed at preserving overall health and quality of life among vulnerable older adults. Abbreviations MCI Mild Cognitive Impairment OF Oral Frailty Declarations Acknowledgement Not applicable. Authors’ contributions YKZ and JJW applied for approval from the ethics committee and designed this study. YKZ, JJW, XML and JL collected clinical data. YKZ performed data analysis and constructed the predictive model. YKZ and JJW validated the models and conducted whole process supervision. JL revised and edited the manuscript. All authors read and approved the final manuscript. Data Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Ethics approval and consent to participate Ethical approval for this study was obtained from the Ethics Committee of the General Hospital of the Northern Theater Command (Approval No: Y (2026) 08). All procedures performed in this study involving human participants were conducted in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Written informed consent was obtained from all participants prior to enrollment in the study. Competing interests The authors declare no conflicts of interest References Ijaz N, Jamil Y, Brown C H t, et al. Role of Cognitive Frailty in Older Adults With Cardiovascular Disease. J Am Heart Assoc 2024;13(4):e033594. Mian M, Tahiri J, Eldin R, et al. Overlooked cases of mild cognitive impairment: Implications to early Alzheimer's disease. Ageing research reviews 2024;98:102335. Yu J, Ng T K S, Mahendran R. Cognitive and physical age gaps in relation to mild cognitive impairment and behavioral phenotypes. GeroScience 2024;46(1):1129-1140. Jia L, Du Y, Chu L, et al. Prevalence, risk factors, and management of dementia and mild cognitive impairment in adults aged 60 years or older in China: a cross-sectional study. The Lancet Public health 2020;5(12):e661-e671. Harwood R H, Goldberg S E, Brand A, et al. Promoting Activity, Independence, and Stability in Early Dementia and mild cognitive impairment (PrAISED): randomised controlled trial. BMJ 2023;382:e074787. Hui Z, Zhao Y, Gong Q, et al. Mediating effect of social participation in the association between cognitive frailty and health-related quality of life among hospitalized older adults: a cross-sectional study in Xi'an, China. BMC Geriatr 2025;25(1):264. Tonetti M S, Bottenberg P, Conrads G, et al. Dental caries and periodontal diseases in the ageing population: call to action to protect and enhance oral health and well-being as an essential component of healthy ageing - Consensus report of group 4 of the joint EFP/ORCA workshop on the boundaries between caries and periodontal diseases. J Clin Periodontol 2017;44 Suppl 18:S135-s144. Dibello V, Zupo R, Sardone R, et al. Oral frailty and its determinants in older age: a systematic review. The lancet Healthy longevity 2021;2(8):e507-e520. Xia S, Wu Y, Wang Q, et al. Oral frailty and nutritional appetite in older adults: a cross-sectional, moderated mediation model of oral self-efficacy and personal mastery. BMC Oral Health 2025;25(1):1503. Dibello V, Solfrizzi V, Lozupone M, et al. Targeting oral frailty indicators of late-life cognitive disorders and depression: a systematic review. Age Ageing 2025;54(7). Fei Y, Niu S, Xi X, et al. Physical frailty intensifies the positive association of oral frailty with poor global cognitive function and executive function among older adults especially for females: a cross-sectional study. BMC Geriatr 2024;24(1):468. Qiao F, Guo H, Zhang J, et al. Association Between Number of Missing Teeth and Hyperlipidemia: The TCLSIH Cohort Study. Journal of inflammation research 2024;17:1095-1104. Chalmers J C, Hernandez-Kapila Y L. The role of the oral microbiome, host response, and periodontal disease treatment in Alzheimer's disease: A primer. Periodontol 2000 2025;98(1):220-227. Dibello V, Custodero C, Cavalcanti R, et al. Impact of periodontal disease on cognitive disorders, dementia, and depression: a systematic review and meta-analysis. GeroScience 2024;46(5):5133-5169. Nagatani M, Tanaka T, Son B K, et al. Oral frailty as a risk factor for mild cognitive impairment in community-dwelling older adults: Kashiwa study. Exp Gerontol 2023;172:112075. Yang M, Liu Y, Watanabe Miura K, et al. Frailty Risk Patterns and Mortality Prediction in Community-Dwelling Older Adults: A 3-Year Longitudinal Study. J Am Med Dir Assoc 2025;26(1):105359. Fujita M, Takeuchi K, Tamada Y, et al. Socioeconomic Inequalities in Oral Frailty Among Older Adults: A Causal Mediation Analysis on the Role of Prevention From Tooth Loss. J Oral Rehabil 2025;52(12):2412-2419. Lv K, Yu P, Xue Y, et al. Construction and validation of an oral frailty risk prediction model for community-dwelling older adults. BMC Geriatr 2025;25(1):808. Lv J, Li J, Wang Y, et al. Construction and validation of a risk prediction model for oral frailty in elderly patients with esophageal cancer. Front Oncol 2025;15:1736063. Liu H, Hu X, Liu Q, et al. Development and internal validation of a risk prediction model for oral frailty in hospitalized older adults with chronic diseases. Frontiers in public health 2026;14:1717485. O'Conor R, Russell A M, Pack A, et al. Managing medications among individuals with mild cognitive impairment and dementia: Patient-caregiver perspectives. J Am Geriatr Soc 2024;72(10):3011-3021. Ornish D, Madison C, Kivipelto M, et al. Effects of intensive lifestyle changes on the progression of mild cognitive impairment or early dementia due to Alzheimer's disease: a randomized, controlled clinical trial. Alzheimers Res Ther 2024;16(1):122. Aponte Ribero V, Efthimiou O, Alwan H, et al. Development and validation of the CARE-DM model to predict the cardiovascular risk in older persons with type 2 diabetes. European journal of preventive cardiology 2026;33(1):44-52. Li T, Shen Y, Leng Y, et al. The prevalence of oral frailty among older adults: a systematic review and meta‑analysis. Eur Geriatr Med 2024;15(3):645-655. Albert M S, DeKosky S T, Dickson D, et al. The diagnosis of mild cognitive impairment due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimer's & dementia : the journal of the Alzheimer's Association 2011;7(3):270-279. Tan J P, Li N, Gao J, et al. Optimal cutoff scores for dementia and mild cognitive impairment of the Montreal Cognitive Assessment among elderly and oldest-old Chinese population. J Alzheimers Dis 2015;43(4):1403-1412. Tanaka T, Hirano H, Ohara Y, et al. Oral Frailty Index-8 in the risk assessment of new-onset oral frailty and functional disability among community-dwelling older adults. Arch Gerontol Geriatr 2021;94:104340. Yao R, Yao L, Yuan C, et al. Accuracy of Calf Circumference Measurement, SARC-F Questionnaire, and Ishii's Score for Screening Stroke-Related Sarcopenia. Front Neurol 2022;13:880907. Yesavage J A, Brink T L, Rose T L, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res 1982;17(1):37-49. Rubenstein L Z, Harker J O, Salvà A, et al. Screening for undernutrition in geriatric practice: developing the short-form mini-nutritional assessment (MNA-SF). J Gerontol A Biol Sci Med Sci 2001;56(6):M366-372. Okeh U, Okoro C J J B B. Evaluating measures of indicators of diagnostic test performance: fundamental meanings and formulars.J Biom Biostat 2012;3(1):2. Lin T, Liang R, Song Q, et al. Development and Validation of PRE-SARC (PREdiction of SARCopenia Risk in Community Older Adults) Sarcopenia Prediction Model. J Am Med Dir Assoc 2024;25(9):105128. Chen Y, Zhang L, Yan W, et al. Factors associated with oral frailty in older adults: a systematic review and meta-analysis.Front Public Health 2025;Volume 13 - 2025. Yi Y, Lee C H, Shin H S, et al. Oral diseases as emerging risk factors for Alzheimer's disease: A scoping review. Jpn Dent Sci Rev 2025;61:292-300. Trends in the global, regional, and national burden of oral conditions from 1990 to 2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet 2025;405(10482):897-910. Park Y S, Hong H P, Ryu S R, et al. Effects of textured food masticatory performance in older people with different dental conditions. BMC Geriatr 2022;22(1):384. Yang C, Li X, An R, et al. Oral Frailty Assessment Tool: Instrument Development and Cross-Sectional Validation Study. J Oral Rehabil 2025;52(10):1608-1618. Lee J Y, Chun A Y, Bae Y, et al. Oral Function as a Predictor of Skeletal Muscle Mass Decline in Community-Dwelling Older Adults. J Oral Rehabil 2025;52(12):2259-2268. Lexomboon D, Kumar A, Freyland S, et al. Is poor chewing ability a risk factor for malnutrition? A six-year longitudinal study of older adults in Sweden. J Nutr Health Aging 2025;29(6):100554. Strączek A, Szałkowska J, Sutkowska P, et al. Impact of nutrition on the condition of the oral mucosa and periodontium: A narrative review. Dental and medical problems 2023;60(4):697-707. Ju Y J, Lee J E, Lee S Y. Associations between Chewing Difficulty, Subjective Cognitive Decline, and Related Functional Difficulties among Older People without Dementia: Focus on Body Mass Index. J Nutr Health Aging 2021;25(3):347-355. Sun Y, Liu H, Li X, et al. Association of social frailty, sarcopenia, and oral frailty with depressive symptoms in Chinese older adults: a cross-sectional study. BMC Public Health 2025;25(1):464. Kawamura K, Maeda K, Miyahara S, et al. Oral hypofunction and social aspects in older adults visiting frailty outpatient clinic. J Oral Rehabil 2024;51(10):2150-2157. Tables Table 1 Baseline Characteristics of MCI Patients According to Oral Frailty Status Characteristics Total (456) Non-Oral Frailty(119) Oral Frailty (337) x 2 / Z value P value Sex 0.026 0.871 male 250(54.8) 66(55.5) 184(54.6) female 206(45.2) 53(44.5) 153(45.4) Age,year,median(IQR) 68.00(63.00,72.00) 66.00(62.00,70.00) 68.00(64.00,72.00) -3.612 < 0.001 Age group 9.012 0.011 60~69 year 275(60.3) 85(71.4) 190(56.4) 70~79 year 156(34.2) 31(26.1) 125(37.1) ≥80 year 25(5.5) 3(2.5) 22(6.5) BMI,kg/m 2 2.303 0.316 18.5~23.9 137(30.0) 37(31.0) 100(29.7) 24.0~27.9 182(40.0) 41(34.5) 141(41.8) ≥28 137(30.0) 41(34.5) 96(28.5) Marriage 4.176 0.124 Married 422(92.5) 115(96.6) 307(91.1) Divorced 14(3.1) 1(0.8) 13(3.9) Widowed 20(4.4) 3(2.5) 17(5.0) Education Level 9.553 0.049 Elementary school or lower 123(27.0) 23(19.3) 100(29.7) Junior high school 205(45.0) 55(46.2) 150(44.5) High school 86(18.9) 32(26.9) 54(16.0) Associate's degree 27(5.9) 6(5.0) 21(6.2) Bachelor's degree or higher 15(3.2) 3(2.5) 12(3.6) Income 4.639 0.200 ＜1000 RMB 65(14.3) 11(9.2) 54(16.0) 1000~2999 RMB 199(43.6) 50(42.1) 149(44.2) 3000~4999 RMB 141(30.9) 43(36.1) 98(29.1) ＞5000 RMB 51(11.2) 15(12.6) 36(10.7) Medical insurance 5.222 0.265 Residents' Medical 304(66.7) 84(70.7) 220(65.3) New Rural Cooperative 116(25.4) 26(21.8) 90(26.7) Subsistence Allowance 15(3.3) 5(4.2) 10(3.0) Commercial Insurance 7(1.5) 3(2.5) 4(1.2) Self-pay 14(3.1) 1(0.8) 13(3.9) Residential Setting 4.972 0.026 Towns 279(61.2) 83(69.7) 196(58.2) Rural 177(38.8) 36(30.3) 141(41.8) Living Arrangement 9.858 0.020 With spouse and children 47(10.3) 8(6.7) 39(11.6) With spouse 317(69.5) 96(80.7) 221(65.6) With children 51(11.2) 7(5.9) 44(13.1) Live alone 41(9.0) 8(6.7) 33(9.8) Smoking 16.224 < 0.001 Yes 156(34.2) 25(21.0) 131(38.9) No 274(60.1) 90(75.6) 184(54.6) Quit smoking 26(5.7) 4(3.4) 22(6.5) Drinking 3.135 0.209 Yes 137(30.0) 32(26.9) 105(31.2) No 292(64.0) 83(69.7) 209(62.0) Quit drinking 27(5.9) 4(3.4) 23(6.8) Comorbidity 19.300 < 0.001 0 type 87(19.1) 37(31.1) 50(14.8) 1~3 type 332(72.8) 79(66.4) 253(75.1) ＞3 type 37(8.1) 3(2.5) 34(10.1) Sarcopenia,median(IQR) 1.00(0.00,3.00) 1.00(0.00,1.00) 2.00(0.00,4.00) -5.326 < 0.001 18.659 < 0.001 Yes 113(24.8) 12(10.1) 101(30.0) No 343(75.2) 107(89.9) 236(70.0) Malnutrition,median(IQR) 10.00(8.00,11.0) 11.00(10.00,11.00) 9.00(8.00,11.00) -5.514 < 0.001 40.905 < 0.001 Yes 263(57.7) 39(32.8) 224(66.5) No 193(42.3) 80(67.2) 113(33.5) Chewing Difficulties 72.311 < 0.001 Yes 248(54.4) 25(21.0) 223(66.2) No 208(45.6) 94(79.0) 114(33.8) Denture 218.451 < 0.001 None 181(39.7) 115(96.6) 66(19.6) 1~5 dentures 236(51.8) 2(1.7) 234(69.4) ＞5 dentures 39(8.6) 2(1.7) 37(11.0) Depression,median(IQR) 9.00(7.00,10.00) 9.00(8.00,10.00) 8.00(7.00,9.00) -4.333 < 0.001 Table 2 The parameters of the three types of machine learning prediction models Model AUC(95% CI) Sensitivity Specificity Accuracy F1 score Binary Logistic Regression 0.917(0.855,0.979) 0.911 0.964 0.957 0.928 Decision Tree 0.907(0.852,0.961) 0.861 0.889 0.956 0.906 XGBoost 0.917(0.857,0.976) 0.821 0.944 0.976 0.892 Table 3 The logistic regression analysis of oral frailty Variable β SE z P OR 95% CI Living Arrangement -3.057 1.303 -2.346 0.019 0.047 0.003~0.545 Malnutrition 2.089 0.550 3.795 < 0.001 8.076 2.909~25.738 Chewing Difficulty 2.489 0.532 4.681 < 0.001 12.055 4.483~36.770 Denture 5.082 0.770 6.596 < 0.001 161.016 42.603~920.001 Depression -0.407 0.155 -2.631 0.008 0.665 0.484~0.893 Additional Declarations No competing interests reported. Supplementary Files Supplementary.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers invited by journal 25 Mar, 2026 Editor assigned by journal 23 Mar, 2026 Editor invited by journal 27 Feb, 2026 Submission checks completed at journal 26 Feb, 2026 First submitted to journal 26 Feb, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8962009","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":612413523,"identity":"e3daca7c-01ce-427f-a0d7-226ef57c2891","order_by":0,"name":"Yangkun Zhang","email":"","orcid":"","institution":"General Hospital of Shenyang Military Command: General Hospital of Northern Theatre command","correspondingAuthor":false,"prefix":"","firstName":"Yangkun","middleName":"","lastName":"Zhang","suffix":""},{"id":612413524,"identity":"49b90d9f-65c5-4e14-b3f1-ee2451cd9432","order_by":1,"name":"Jingjing Wang","email":"","orcid":"","institution":"General Hospital of Shenyang Military Command: General Hospital of Northern Theatre command","correspondingAuthor":false,"prefix":"","firstName":"Jingjing","middleName":"","lastName":"Wang","suffix":""},{"id":612413525,"identity":"91dc4fe2-0210-4403-9248-4b8eea3adbab","order_by":2,"name":"Xiaoming Liu","email":"","orcid":"","institution":"General Hospital of Shenyang Military Command: General Hospital of Northern Theatre command","correspondingAuthor":false,"prefix":"","firstName":"Xiaoming","middleName":"","lastName":"Liu","suffix":""},{"id":612413527,"identity":"afeb9d33-4a05-4bf8-a4c6-54c70432238b","order_by":3,"name":"Jing Li","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1klEQVRIie3RsQrCMBCA4UggLiddT5T4BMJBQCyIvkpDVx/ASSqugqtCH0IQnFMCuvgABScRnFycHBRRN3Fp3ATz7x93xzHm8/1gQcNmhxugFJwbN1KdiZig3lZBWURuhHJoIXQGej4FctxslxBiHxVZOOcn1pXNpECUUhMRbVGSrSzDlMWqZQoIZ5GJ9OQ1pbKqATN6VUQE04nJ7qgXFo5uBDAujRJAPR+DcCMIa86fUAVcqDAlh1t6m+nlymAoRWD3+WnQlYXkY6Tra97It8Ln8/n+ogfJXT1+fTao1gAAAABJRU5ErkJggg==","orcid":"","institution":"General Hospital of Shenyang Military Command: General Hospital of Northern Theatre command","correspondingAuthor":true,"prefix":"","firstName":"Jing","middleName":"","lastName":"Li","suffix":""}],"badges":[],"createdAt":"2026-02-25 01:53:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8962009/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8962009/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":105641147,"identity":"4451602c-e1cb-4f60-bb09-27aa2620ffdf","added_by":"auto","created_at":"2026-03-28 16:26:07","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":140262,"visible":true,"origin":"","legend":"\u003cp\u003eRandom Forest Variable Importance Ranking\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8962009/v1/06a3f11896830d46181660e9.png"},{"id":105728930,"identity":"c41b4844-2bf5-42ca-8b8f-ae1efb1e7f3c","added_by":"auto","created_at":"2026-03-30 11:13:02","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":96584,"visible":true,"origin":"","legend":"\u003cp\u003e(a) Cross validation bias plot of Lasso logistic regression\u003c/p\u003e\n\u003cp\u003e(b) Coefficient Path Plot of Lasso logistic regression\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8962009/v1/98b4c15c9ac70f7bc7b0073f.png"},{"id":105641152,"identity":"ad20877a-5d91-48c2-a680-419d49eeb498","added_by":"auto","created_at":"2026-03-28 16:26:07","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":73202,"visible":true,"origin":"","legend":"\u003cp\u003eROC curve for the test set\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-8962009/v1/ed5071c1fcb2f9598235e169.png"},{"id":105641150,"identity":"b465b535-de7e-47e3-98e1-6f196fd2ce77","added_by":"auto","created_at":"2026-03-28 16:26:07","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":105760,"visible":true,"origin":"","legend":"\u003cp\u003eCalibration curve for the test set\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-8962009/v1/7773255400f78603452535fd.png"},{"id":105641148,"identity":"add670f9-0cbb-49bb-943b-b0a301cf11ad","added_by":"auto","created_at":"2026-03-28 16:26:07","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":99648,"visible":true,"origin":"","legend":"\u003cp\u003eDCA curve for the test set\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-8962009/v1/bac903a808a03611521e96aa.png"},{"id":105752005,"identity":"9b7fbed9-173b-4f1d-90c5-f415b8a22a37","added_by":"auto","created_at":"2026-03-30 15:53:06","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1395886,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8962009/v1/7defd3de-31b2-472c-94d4-a125ae4f6611.pdf"},{"id":105641151,"identity":"8b509b85-b353-4296-8576-ff179b7257cd","added_by":"auto","created_at":"2026-03-28 16:26:07","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":709457,"visible":true,"origin":"","legend":"","description":"","filename":"Supplementary.docx","url":"https://assets-eu.researchsquare.com/files/rs-8962009/v1/c035716645497b72ef96d35a.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Development and Validation of a Risk Prediction Model for Oral Frailty in Older Adults With Mild Cognitive Impairment: A Cross-Sectional Study","fulltext":[{"header":"Introduction","content":"\u003cp\u003eWith the accelerating process of global population aging, the coexistence of geriatric syndromes has emerged as a major public health challenge. \u003csup\u003e1\u003c/sup\u003eMild cognitive impairment (MCI), a transitional clinical stage between normal cognitive aging and dementia, is characterized by memory decline and executive dysfunction.\u003csup\u003e2\u003c/sup\u003eIts prevalence increases substantially with advancing age.\u003csup\u003e3\u003c/sup\u003e By 2020, an estimated 38.77 million older adults aged 60 years and above in China were affected by MCI.\u0026nbsp;\u003csup\u003e4\u003c/sup\u003eBeyond its well-established role as a precursor to dementia, MCI is frequently accompanied by systemic functional decline, significantly compromising independence and quality of life.\u003csup\u003e5, 6\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eOral health has increasingly been recognized as a fundamental component of overall health in older adults.\u003csup\u003e7\u003c/sup\u003e Oral frailty is an age-related syndrome involving progressive deterioration in oral structure and function, manifesting as chewing difficulty, swallowing dysfunction, and tooth loss.\u0026nbsp;\u003csup\u003e8\u003c/sup\u003eOral frailty not only impairs nutritional intake and daily functioning but may also interact bidirectionally with cognitive decline.\u003csup\u003e9, 10\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eOlder adults with cognitive impairment often experience diminished memory, executive function, and fine motor coordination, which reduce their ability to maintain adequate oral hygiene and increase susceptibility to dental caries, periodontal disease, and tooth loss.\u0026nbsp;\u003csup\u003e11\u003c/sup\u003eConversely, chronic oral inflammation may serve as a persistent source of systemic low-grade inflammation.\u003csup\u003e12\u003c/sup\u003e Pro-inflammatory cytokines such as tumor necrosis factor-α and interleukin-6 can cross the blood–brain barrier, promote neuroinflammation, and accelerate β-amyloid deposition, thereby exacerbating cognitive deterioration.\u0026nbsp;\u003csup\u003e13, 14\u003c/sup\u003eThis bidirectional interaction suggests that the coexistence of MCI and oral frailty may generate synergistic adverse effects, increasing the risk of falls, malnutrition, disability, and mortality, and imposing substantial burdens on families and healthcare systems.\u003csup\u003e8, 15, 16\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eEmerging evidence indicates that older adults with MCI are at significantly higher risk of developing oral frailty compared with cognitively intact peers.\u003csup\u003e17\u003c/sup\u003e When these conditions coexist, the risk of adverse health outcomes is markedly amplified, presenting complex challenges for geriatric care.\u003c/p\u003e\n\u003cp\u003eDespite this growing recognition, current prediction models for oral frailty have primarily focused on community-dwelling older populations or older adults with chronic disease, and have not specifically addressed the unique characteristics of individuals with MCI.\u003csup\u003e18\u003c/sup\u003e\u003csup\u003e,\u0026nbsp;\u003c/sup\u003e\u003csup\u003e19, 20\u003c/sup\u003eCognitive impairment may obscure symptom reporting, reduce self-care capacity, and alter health behaviors, thereby necessitating targeted risk stratification tools for this vulnerable group.\u003csup\u003e21\u003c/sup\u003e Moreover, clinical management of MCI has traditionally emphasized delaying progression to dementia, whereas systematic assessment and integrated intervention for oral health remain underexplored.\u003csup\u003e22\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eMachine learning techniques have demonstrated strong capability in handling multidimensional clinical data and identifying complex nonlinear associations. These methods have achieved promising performance in risk prediction for cardiovascular disease, diabetes, and other chronic conditions.\u0026nbsp;\u003csup\u003e23\u003c/sup\u003eHowever, studies applying machine learning to predict oral frailty specifically among older adults with MCI remain limited.\u003c/p\u003e\n\u003cp\u003eTherefore, this study aimed to develop and internally validate a risk prediction model for oral frailty in older adults with MCI using multiple machine learning algorithms. By integrating multidimensional clinical indicators, we sought to enable early identification of high-risk individuals and provide a foundation for personalized, proactive oral health interventions. Ultimately, this work may contribute to the development of an integrated cognitive–oral health management framework and improve health outcomes in this vulnerable population.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eStudy Design\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis was a prospective observational study conducted to develop and internally validate a clinical prediction model for oral frailty among older adults with MCI.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eParticipants\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eParticipants were recruited using convenience sampling from the Geriatric Outpatient Department of General Hospital of Northern Theater Command\u0026nbsp;between October 2025 and January 2026. Participants were eligible if they: (1) were aged ≥60 years; (2)met diagnostic criteria for mild cognitive impairment; (3)had stable physical conditions and were able to complete questionnaires and oral examinations; (4)provided written informed consent. Participants were excluded if they: (1)had severe psychiatric disorders impairing cooperation; (2)had oral and maxillofacial deformities or tumors; (3)had end-stage systemic diseases (eg, advanced hepatic or renal failure, malignancy).\u003c/p\u003e\n\u003cp\u003eSample size was estimated according to the TRIPOD recommendations for prediction model development.Based on the events-per-variable (EPV) principle, at least 10 outcome events per predictor variable were required to ensure model stability. Assuming an estimated oral frailty prevalence of 57% from prior literature and considering up to 24 candidate predictors, the minimum required sample size was calculated as 421 participants.\u003csup\u003e24\u003c/sup\u003eAfter accounting for an anticipated 5% attrition rate, the final target sample size was set at 442 participants.\u003c/p\u003e\n\u003cp\u003eEthical approval was obtained from the Ethics Committee of\u0026nbsp;General Hospital of Northern Theater Command [ Approval No: Y (2026) 08 ].Written informed consent was obtained from all participants prior to enrollment in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMeasurements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSociodemographic and Clinical Data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA structured questionnaire developed based on literature review and expert consultation was used to collect demographic and clinical information.Collected variables included:sex, age, body mass index (BMI), marriage, education level,income, medical insurance, residential setting, living arrangement, smoking, drinking, comorbidity,dentures.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment of Cognition\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn the present study, mild cognitive impairment (MCI) was diagnosed according to the criteria established by the Alzheimer’s Association McKhann and Albert 2011 criteria.\u003csup\u003e25\u003c/sup\u003e The diagnostic criteria were as follows: (1) for participants aged 60~79 years, 80~89 years, and ≥90 years, cognitive function was assessed using the Peking Union Medical College Hospital version of the Montreal Cognitive Assessment (MoCA-P).\u003csup\u003e26\u003c/sup\u003e The optimal cutoff values for identifying mild cognitive impairment were ≤25, ≤24, and ≤23, respectively, across these three age groups; (2) possible mild impairment in complex instrumental activities of daily living, while basic activities of daily living remained functionally independent; (3) absence of a diagnosis of dementia; and (4) confirmation of the diagnosis by a senior attending neurologist at a tertiary hospital.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment of Oral Frailty\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOral frailty among older adults with MCI was assessed using the Oral Frailty\u0026nbsp;Index-8 Scale (OFI-8) developed by Tanaka et al.\u003csup\u003e27\u003c/sup\u003e The scale comprises five domains: denture use, swallowing function, social participation, oral health behaviors, and masticatory function, encompassing a total of eight items. Each item is answered dichotomously (“yes” or “no”). The first three items are scored from 0 to 2 points, whereas the remaining five items are scored from 0 to 1 point, yielding a total score ranging from 0 to 11. A total score ≥4 indicates oral frailty, a score of 3 denotes prefrailty, and a score of 0~2 indicates no oral frailty.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment of Sarcopenia\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSarcopenia status was evaluated using the\u0026nbsp;Sarcopenia-Five (SARC-F) questionnaire developed by Saint Louis University.\u003csup\u003e28\u003c/sup\u003e The scale consists of five items assessing strength, assistance in walking, rising from a chair, climbing stairs, and falls. Each item is scored from 0 to 2 points, with a total score ranging from 0 to 10. A total score ≥4 indicates a high risk of sarcopenia.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment of Depression\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDepressive symptoms were assessed using the 15-item Geriatric Depression Scale (GDS-15) developed by Yesavage et al.\u003csup\u003e29\u003c/sup\u003e The scale comprises two dimensions and includes 15 dichotomous (“yes” or “no”) items. Total scores range from 0 to 15, with higher scores indicating more severe depressive symptoms. A score ≥5 is considered indicative of clinically significant depressive symptoms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAssessment of Malnutrition\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNutritional status was assessed using the Mini Nutritional Assessment–Short Form (MNA-SF) developed by Rubenstein et al.\u003csup\u003e30\u003c/sup\u003e The instrument includes six items evaluating BMI, recent weight loss, medical history, mobility, neuropsychological status, and dietary intake. Total scores range from 0 to 14, with scores ≥11 indicating normal nutritional status and scores \u0026lt;11 indicating risk of malnutrition or malnutrition.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Collection Procedures\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eData were collected through face-to-face interviews and clinical examinations. A multidisciplinary research team consisting of geriatricians, dental physicians, geriatric nurses, and public health experts underwent standardized training before study initiation. Eligible participants were screened according to inclusion and exclusion criteria. After obtaining informed consent, trained researchers conducted one-on-one structured interviews and clinical oral examinations. All questionnaires were reviewed on-site to ensure completeness. Oral examinations and sarcopenia assessments were performed according to standardized clinical procedures.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStatistical Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll statistical analyses were conducted using R software (version 4.5.2, R Foundation for Statistical Computing). A two-sided P value \u0026lt;0.05 was considered statistically significant.Continuous variables were assessed for normality using the Shapiro–Wilk test. As most continuous variables were non-normally distributed, they were expressed as median and interquartile range (IQR) and compared using the Mann–Whitney U test. Categorical variables were presented as frequencies and percentages and compared using the chi-square test or Fisher’s exact test as appropriate. No missing data were observed in the variables included in the analysis.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel Development Strategy\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe primary objective of this study was to develop and internally validate a prediction model for oral frailty among older adults with MCI. Stratified random sampling based on outcome status was applied to maintain similar event proportions across datasets.The entire cohort was randomly divided into a training dataset (70%) for model development and a validation dataset (30%) for independent performance evaluation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFeature Selection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA two-stage feature selection process was implemented to reduce dimensionality and minimize overfitting.First, variables demonstrating statistical significance in univariate analyses were entered into a random forest model to estimate relative variable importance.Second, least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation was performed to identify the most predictive subset of variables. The lambda.1se criterion was adopted to select a parsimonious model while maintaining stability.Variance inflation factors (VIFs) were examined to assess multicollinearity among retained predictors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel Construction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eBased on the selected predictors, three classification models were developed:Binary logistic regression; Decision tree; Extreme gradient boosting (XGBoost). Model development was conducted in the training dataset. Model performance was primarily evaluated in the independent validation dataset.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel Performance Evaluation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eModel discrimination was assessed using the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and accuracy were calculated based on predicted probabilities.\u003csup\u003e31\u003c/sup\u003e The optimal probability threshold was determined using the Youden index. Model calibration was evaluated using:Hosmer–Lemeshow goodness-of-fit test,calibration plots,calibration slope and intercept.\u003csup\u003e32\u003c/sup\u003eThese metrics assessed agreement between predicted probabilities and observed outcomes.Overall predictive accuracy was evaluated using the Brier score, which measures the mean squared difference between predicted probabilities and observed outcomes.Internal validation was performed using bootstrap resampling (1,000 repetitions) in the training dataset to estimate optimism-corrected performance measures.The events-per-variable (EPV) ratio was calculated to assess model stability.Decision curve analysis (DCA) was performed to evaluate the net clinical benefit of the final model across a range of threshold probabilities.\u003csup\u003e32\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSensitivity Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo assess the influence of dentures on model performance, a sensitivity analysis was conducted by excluding dentures from the model. The resulting AUC was compared with that of the primary model.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eParticipant Characteristics\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 500 older adults who met the diagnostic criteria for mild cognitive impairment (MCI) were screened. After excluding 28 individuals who declined participation and 16 who withdrew during follow-up, 456 participants were included in the final analysis. Among them, 337 were identified as having oral frailty according to the OFI-8 criteria, corresponding to a prevalence of 73.9%.\u003c/p\u003e\n\u003cp\u003eThe median age of participants was 68.0 years (interquartile range [IQR], 63.0–72.0), and 54.8% were male. The cohort was randomly divided into a training set (n = 319, 236 events) and an independent validation set (n = 137, 101 events). No statistically significant differences in baseline characteristics were observed between the 2 datasets (Supplementary Table 1), indicating appropriate distributional balance.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eUnivariate analysis of oral frailty\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eContinuous variables were standardized ; categorical variables were dummy-coded. Between-group comparisons revealed significant differences in age, educational level, residential setting, living arrangement, smoking status, comorbidities, sarcopenia, malnutrition, chewing difficulty, denture, and depressive symptoms (all P\u0026nbsp;\u0026lt; 0.05). Detailed results are presented in Table 1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eVariable Selection\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA two-stage feature selection strategy was applied. First, variables with statistical significance in univariate analyses were entered into a random forest model to estimate relative feature importance. Subsequently, the top-ranked predictors were subjected to least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation. The lambda.1se criterion was adopted to obtain a parsimonious model while minimizing overfitting.\u003c/p\u003e\n\u003cp\u003eNine predictors were retained for model construction: denture, chewing difficulty, sarcopenia, malnutrition, depressive symptoms, educational level, smoking status, comorbidities, and living arrangement.As shown in Figure 1 and Figure 2. Multicollinearity among retained predictors was assessed using variance inflation factors (VIFs). All VIF values were below 2, indicating no evidence of significant multicollinearity.Pairwise correlations among predictors were low, absolute Spearman correlation coefficients \u0026lt; 0.30 (Supplementary Figure1).\u003c/p\u003e\n\u003cp\u003eThe events-per-variable ratio exceeded 30 (337 events / 9 predictors), indicating adequate model stability.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eModel Comparison\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThree classification models were developed: binary logistic regression, decision tree, and extreme gradient boosting (XGBoost). Model performance was primarily evaluated in the validation dataset.The comparative performance results are summarized in Table 2.Although the three classification models demonstrated comparable discrimination in the validation dataset, the binary logistic regression model was selected as the final prediction model due to its balanced sensitivity, interpretability, and clinical applicability. While XGBoost showed slightly higher specificity, logistic regression provided superior sensitivity and comparable AUC performance, supporting its use as the primary model.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFinal Model Performance\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDiscrimination\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn the validation dataset, the final logistic regression model achieved an AUC of 0.917 (95% CI: 0.855–0.979), indicating excellent discrimination. The corresponding training AUC was 0.965 (95% CI: 0.946–0.984). The ROC curve of the validation set is shown in Figure 3.These findings suggest strong ability to distinguish patients with and without oral frailty.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCalibration\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCalibration was assessed using the Hosmer–Lemeshow goodness-of-fit test and calibration plots. The Hosmer–Lemeshow test indicated no evidence of poor fit (\u003cem\u003ex\u003c/em\u003e² = 6.44, df = 8, P = 0.598). The calibration slope and intercept suggested acceptable agreement between predicted and observed probabilities. Calibration curves confirmed adequate model calibration(Figure 4).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInternal Validation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eInternal validation using 1,000 bootstrap resamples yielded an optimism-corrected AUC of 0.945 in the training dataset, suggesting minimal overfitting.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOverall Accuracy\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe Brier score was 0.057 in the training set and 0.100 in the validation set, indicating good overall predictive accuracy.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical Utility\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDecision curve analysis demonstrated that the final model provided greater net benefit across a broad range of threshold probabilities compared with treat-all and treat-none strategies, supporting its potential clinical applicability(Figure 5).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eOptimal Threshold\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study determined the optimal threshold using the Youden index rather than the conventional 0.5 cutoff.The optimal probability threshold determined using the Youden index was 0.696, corresponding to a sensitivity of 0.911 and specificity of 0.964 in the validation dataset.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLogistic Regression Estimates\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMultivariable logistic regression identified denture, chewing difficulty, malnutrition, living arrangement, and depressive symptoms as independent predictors of oral frailty (Table 3). Denture demonstrated the strongest association with oral frailty (OR 161.02; 95% CI, 42.60~920.00; P \u0026lt; 0.001).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSensitivity Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo assess the contribution of denture number to overall model performance, a sensitivity analysis excluding this variable was conducted. In the validation dataset, the AUC decreased to 0.797(Supplementary Figure 2), indicating that denture burden substantially contributes to the model’s discriminative ability.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this prospective cohort of older adults with mild cognitive impairment (MCI), we developed and internally validated a multidimensional risk prediction model for oral frailty that demonstrated robust discrimination and calibration performance. The identification of dentures, chewing difficulty, malnutrition, depressive symptoms, and living arrangement as influential predictors reinforces the concept of oral frailty as a systemic geriatric syndrome rather than an isolated oral disorder. Our findings highlight the intertwined structural, functional, nutritional, and psychosocial determinants of oral vulnerability in cognitively impaired populations and provide a clinically pragmatic framework for early risk stratification.\u003c/p\u003e\n\u003cp\u003eIn this study, the prevalence of oral frailty among older adults with mild cognitive impairment was 73.9%, underscoring the complex and potentially bidirectional relationship between oral frailty and cognitive impairment in later life. Population-based longitudinal data have shown that oral frailty significantly predicts the risk of new-onset MCI among community-dwelling older individuals, even after adjustment for confounding factors such as physical frailty and demographic characteristics.\u003csup\u003e15\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eBeyond incidence associations, the pathophysiological mechanisms connecting oral frailty and cognitive deterioration are increasingly supported by epidemiological and mechanistic studies. Declining oral function may reduce masticatory stimulation and sensory input that support neurocognitive integrity, while chronic oral inflammation can contribute to systemic inflammation and neuroinflammatory cascades that accelerate cognitive impairment.\u003csup\u003e33\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eClinicians should consider early screening using OFI-8 in combination with the MoCA to facilitate timely identification of individuals at elevated risk. Integration of salivary and blood-based biomarkers may further enhance risk stratification and support precision identification of high-risk populations.\u003csup\u003e8\u003c/sup\u003e Interventions should prioritize control of inflammatory burden and functional rehabilitation. Comprehensive periodontal therapy, timely prosthetic rehabilitation for tooth loss, and standardized management of xerostomia are essential to reduce chronic oral infection and inflammation. \u003csup\u003e34\u003c/sup\u003e These measures should be complemented by masticatory function training and nutritional support—particularly adequate protein intake—to mitigate the development of oral frailty and potentially slow the progression of cognitive decline in older adults.\u003c/p\u003e\n\u003cp\u003eIn the context of clinical assessment, our selection of denture number as a graded indicator rather than a binary variable reflects the clinical reality that severity of tooth loss and prosthetic burden may more accurately represent oral function decline.\u003csup\u003e35\u003c/sup\u003e The number of dentures better captures the magnitude of oral functional loss and was highly predictive in our model.Although the magnitude of the estimated odds ratio was substantial, this likely reflects the categorical structure of the variable and distribution imbalance between exposure groups rather than a literal 160-fold biological increase in risk. Extensive tooth loss and reliance on multiple dentures may serve as markers of cumulative structural oral deterioration, compromised occlusal stability, and reduced masticatory efficiency.\u003csup\u003e36\u003c/sup\u003e However, the sensitivity analysis excluding denture number demonstrated reduced discrimination (AUC = 0.797), indicating that although denture burden carries strong predictive signal, it should be interpreted in the context of comprehensive oral function rather than as a sole defining criterion. This aligns with emerging views that multidimensional indicators of oral status, such as chewing difficulty, swallowing impairment, and salivary function, may offer more nuanced insights into overall oral frailty and its relationship with systemic outcomes.\u003csup\u003e37\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eNotably, several risk factors identified in our model, including chewing difficulty, malnutrition, depressive symptoms, and living arrangement have also been implicated as determinants of both oral frailty and cognitive dysfunction.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eMastication represents a critical functional link between oral structures and systemic health. \u003csup\u003e38\u003c/sup\u003eChewing difficulty\u0026nbsp;may reduce dietary diversity and protein intake, thereby increasing the risk of malnutrition.\u003csup\u003e39\u003c/sup\u003e In turn, malnutrition can compromise oral mucosal integrity, delay wound healing, and weaken the muscular strength required for effective swallowing.\u003csup\u003e40\u003c/sup\u003e\u003c/p\u003e\n\u003cp\u003eGiven the limited adherence to oral self-care commonly observed in individuals with MCI, a caregiver-led oral care model should be considered in clinical practice. Standardized chewing rehabilitation programs aimed at improving muscle strength and coordination may help prevent further functional decline. In addition, individualized dietary plans tailored to patients’ graded chewing capacity are recommended to avoid monotonous diets and nutritional deficiencies resulting from chewing limitations. Such comprehensive strategies may help delay the progression of sarcopenia and physical frailty and potentially mitigate downstream contributors to cognitive decline.\u003csup\u003e41\u003c/sup\u003e Depressive symptoms were significantly associated with oral frailty, suggesting that oral frailty should not be regarded solely as a functional oral condition. Rather, it may represent a broader geriatric frailty phenotype that encompasses emotional and psychosocial dimensions.\u003csup\u003e42\u003c/sup\u003e Living arrangement emerged as an independent predictor of oral frailty, suggesting that the social environment may substantially influence oral health behaviors, dietary patterns, and access to care.\u003csup\u003e43\u003c/sup\u003e Older adults with MCI who live with a spouse or adult children may benefit from shared meals, informal health monitoring, and caregiving support, thereby reducing the risk of developing oral frailty.\u003c/p\u003e\n\u003cp\u003eOur study adds to the growing literature advocating for integrated predictive tools in geriatric practice. Whereas most existing oral frailty research has focused on prevalence and cross-sectional associations, there remains a paucity of clinical tools specifically tailored to populations with cognitive impairment. Prior models developed for general older adult populations may not account for the unique behavioral and functional characteristics of individuals with MCI, such as reduced self-care capacity and atypical symptom reporting. Our model addresses this gap by leveraging machine learning-assisted variable selection to capture multidimensional relevance while maintaining interpretability through logistic regression—an approach well-aligned with clinical decision-making needs.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLimitations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSeveral limitations merit consideration. First, the single-center design and the lack of external validation may limit generalizability. Future multicenter studies with external cohorts are needed to confirm the transportability of our model. Second, although our model demonstrated promising performance metrics, the OFI-8 instrument primarily relies on self-reported items and screening responses; incorporation of objective oral health measures, such as tongue pressure or periodontal indices, may further enhance predictive utility. Finally, while the prediction model reflects associations rather than causal relationships, it provides clinically actionable risk stratification that may guide targeted oral health and nutritional interventions in older adults with MCI.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eOur study presents a clinically relevant and statistically robust prediction model for oral frailty in older adults with MCI. By integrating clinical, nutritional, and psychosocial indicators, the model facilitates early identification of high-risk individuals, which may inform tailored preventive strategies. Given the growing prevalence of both cognitive impairment and oral frailty in aging populations, such predictive tools hold promise for guiding interdisciplinary interventions aimed at preserving overall health and quality of life among vulnerable older adults.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eMCI \u0026nbsp;Mild Cognitive Impairment\u003c/p\u003e\n\u003cp\u003eOF \u0026nbsp; Oral Frailty\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgement\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors\u0026rsquo; contributions\u003c/strong\u003e \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eYKZ and JJW applied for approval from the ethics committee and designed this study. YKZ, JJW, XML and JL collected clinical data. YKZ performed data analysis and constructed the predictive model. YKZ and JJW validated the models and conducted whole process supervision. JL revised and edited the manuscript. All authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eEthical approval for this study was obtained from the Ethics Committee of the General Hospital of the Northern Theater Command (Approval No: Y (2026) 08). All procedures performed in this study involving human participants were conducted in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Written informed consent was obtained from all participants prior to enrollment in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no conflicts of interest\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eIjaz N, Jamil Y, Brown C H t, et al. Role of Cognitive Frailty in Older Adults With Cardiovascular Disease. J Am Heart Assoc 2024;13(4):e033594.\u003c/li\u003e\n\u003cli\u003eMian M, Tahiri J, Eldin R, et al. Overlooked cases of mild cognitive impairment: Implications to early Alzheimer\u0026apos;s disease. Ageing research reviews 2024;98:102335.\u003c/li\u003e\n\u003cli\u003eYu J, Ng T K S, Mahendran R. Cognitive and physical age gaps in relation to mild cognitive impairment and behavioral phenotypes. GeroScience 2024;46(1):1129-1140.\u003c/li\u003e\n\u003cli\u003eJia L, Du Y, Chu L, et al. Prevalence, risk factors, and management of dementia and mild cognitive impairment in adults aged 60 years or older in China: a cross-sectional study. The Lancet Public health 2020;5(12):e661-e671.\u003c/li\u003e\n\u003cli\u003eHarwood R H, Goldberg S E, Brand A, et al. Promoting Activity, Independence, and Stability in Early Dementia and mild cognitive impairment (PrAISED): randomised controlled trial. BMJ 2023;382:e074787.\u003c/li\u003e\n\u003cli\u003eHui Z, Zhao Y, Gong Q, et al. Mediating effect of social participation in the association between cognitive frailty and health-related quality of life among hospitalized older adults: a cross-sectional study in Xi\u0026apos;an, China. BMC Geriatr 2025;25(1):264.\u003c/li\u003e\n\u003cli\u003eTonetti M S, Bottenberg P, Conrads G, et al. Dental caries and periodontal diseases in the ageing population: call to action to protect and enhance oral health and well-being as an essential component of healthy ageing - Consensus report of group 4 of the joint EFP/ORCA workshop on the boundaries between caries and periodontal diseases. J Clin Periodontol 2017;44 Suppl 18:S135-s144.\u003c/li\u003e\n\u003cli\u003eDibello V, Zupo R, Sardone R, et al. Oral frailty and its determinants in older age: a systematic review. The lancet Healthy longevity 2021;2(8):e507-e520.\u003c/li\u003e\n\u003cli\u003eXia S, Wu Y, Wang Q, et al. Oral frailty and nutritional appetite in older adults: a cross-sectional, moderated mediation model of oral self-efficacy and personal mastery. BMC Oral Health 2025;25(1):1503.\u003c/li\u003e\n\u003cli\u003eDibello V, Solfrizzi V, Lozupone M, et al. Targeting oral frailty indicators of late-life cognitive disorders and depression: a systematic review. Age Ageing 2025;54(7).\u003c/li\u003e\n\u003cli\u003eFei Y, Niu S, Xi X, et al. Physical frailty intensifies the positive association of oral frailty with poor global cognitive function and executive function among older adults especially for females: a cross-sectional study. BMC Geriatr 2024;24(1):468.\u003c/li\u003e\n\u003cli\u003eQiao F, Guo H, Zhang J, et al. Association Between Number of Missing Teeth and Hyperlipidemia: The TCLSIH Cohort Study. Journal of inflammation research 2024;17:1095-1104.\u003c/li\u003e\n\u003cli\u003eChalmers J C, Hernandez-Kapila Y L. The role of the oral microbiome, host response, and periodontal disease treatment in Alzheimer\u0026apos;s disease: A primer. Periodontol 2000 2025;98(1):220-227.\u003c/li\u003e\n\u003cli\u003eDibello V, Custodero C, Cavalcanti R, et al. Impact of periodontal disease on cognitive disorders, dementia, and depression: a systematic review and meta-analysis. GeroScience 2024;46(5):5133-5169.\u003c/li\u003e\n\u003cli\u003eNagatani M, Tanaka T, Son B K, et al. Oral frailty as a risk factor for mild cognitive impairment in community-dwelling older adults: Kashiwa study. Exp Gerontol 2023;172:112075.\u003c/li\u003e\n\u003cli\u003eYang M, Liu Y, Watanabe Miura K, et al. Frailty Risk Patterns and Mortality Prediction in Community-Dwelling Older Adults: A 3-Year Longitudinal Study. J Am Med Dir Assoc 2025;26(1):105359.\u003c/li\u003e\n\u003cli\u003eFujita M, Takeuchi K, Tamada Y, et al. Socioeconomic Inequalities in Oral Frailty Among Older Adults: A Causal Mediation Analysis on the Role of Prevention From Tooth Loss. J Oral Rehabil 2025;52(12):2412-2419.\u003c/li\u003e\n\u003cli\u003eLv K, Yu P, Xue Y, et al. Construction and validation of an oral frailty risk prediction model for community-dwelling older adults. BMC Geriatr 2025;25(1):808.\u003c/li\u003e\n\u003cli\u003eLv J, Li J, Wang Y, et al. Construction and validation of a risk prediction model for oral frailty in elderly patients with esophageal cancer. Front Oncol 2025;15:1736063.\u003c/li\u003e\n\u003cli\u003eLiu H, Hu X, Liu Q, et al. Development and internal validation of a risk prediction model for oral frailty in hospitalized older adults with chronic diseases. Frontiers in public health 2026;14:1717485.\u003c/li\u003e\n\u003cli\u003eO\u0026apos;Conor R, Russell A M, Pack A, et al. Managing medications among individuals with mild cognitive impairment and dementia: Patient-caregiver perspectives. J Am Geriatr Soc 2024;72(10):3011-3021.\u003c/li\u003e\n\u003cli\u003eOrnish D, Madison C, Kivipelto M, et al. Effects of intensive lifestyle changes on the progression of mild cognitive impairment or early dementia due to Alzheimer\u0026apos;s disease: a randomized, controlled clinical trial. Alzheimers Res Ther 2024;16(1):122.\u003c/li\u003e\n\u003cli\u003eAponte Ribero V, Efthimiou O, Alwan H, et al. Development and validation of the CARE-DM model to predict the cardiovascular risk in older persons with type 2 diabetes. European journal of preventive cardiology 2026;33(1):44-52.\u003c/li\u003e\n\u003cli\u003eLi T, Shen Y, Leng Y, et al. The prevalence of oral frailty among older adults: a systematic review and meta‑analysis. Eur Geriatr Med 2024;15(3):645-655.\u003c/li\u003e\n\u003cli\u003eAlbert M S, DeKosky S T, Dickson D, et al. The diagnosis of mild cognitive impairment due to Alzheimer\u0026apos;s disease: recommendations from the National Institute on Aging-Alzheimer\u0026apos;s Association workgroups on diagnostic guidelines for Alzheimer\u0026apos;s disease. Alzheimer\u0026apos;s \u0026amp; dementia : the journal of the Alzheimer\u0026apos;s Association 2011;7(3):270-279.\u003c/li\u003e\n\u003cli\u003eTan J P, Li N, Gao J, et al. Optimal cutoff scores for dementia and mild cognitive impairment of the Montreal Cognitive Assessment among elderly and oldest-old Chinese population. J Alzheimers Dis 2015;43(4):1403-1412.\u003c/li\u003e\n\u003cli\u003eTanaka T, Hirano H, Ohara Y, et al. Oral Frailty Index-8 in the risk assessment of new-onset oral frailty and functional disability among community-dwelling older adults. Arch Gerontol Geriatr 2021;94:104340.\u003c/li\u003e\n\u003cli\u003eYao R, Yao L, Yuan C, et al. Accuracy of Calf Circumference Measurement, SARC-F Questionnaire, and Ishii\u0026apos;s Score for Screening Stroke-Related Sarcopenia. Front Neurol 2022;13:880907.\u003c/li\u003e\n\u003cli\u003eYesavage J A, Brink T L, Rose T L, et al. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res 1982;17(1):37-49.\u003c/li\u003e\n\u003cli\u003eRubenstein L Z, Harker J O, Salv\u0026agrave; A, et al. Screening for undernutrition in geriatric practice: developing the short-form mini-nutritional assessment (MNA-SF). J Gerontol A Biol Sci Med Sci 2001;56(6):M366-372.\u003c/li\u003e\n\u003cli\u003eOkeh U, Okoro C J J B B. Evaluating measures of indicators of diagnostic test performance: fundamental meanings and formulars.J Biom Biostat 2012;3(1):2.\u003c/li\u003e\n\u003cli\u003eLin T, Liang R, Song Q, et al. Development and Validation of PRE-SARC (PREdiction of SARCopenia Risk in Community Older Adults) Sarcopenia Prediction Model. J Am Med Dir Assoc 2024;25(9):105128.\u003c/li\u003e\n\u003cli\u003eChen Y, Zhang L, Yan W, et al. Factors associated with oral frailty in older adults: a systematic review and meta-analysis.Front Public Health 2025;Volume 13 - 2025.\u003c/li\u003e\n\u003cli\u003eYi Y, Lee C H, Shin H S, et al. Oral diseases as emerging risk factors for Alzheimer\u0026apos;s disease: A scoping review. Jpn Dent Sci Rev 2025;61:292-300.\u003c/li\u003e\n\u003cli\u003eTrends in the global, regional, and national burden of oral conditions from 1990 to 2021: a systematic analysis for the Global Burden of Disease Study 2021. Lancet 2025;405(10482):897-910.\u003c/li\u003e\n\u003cli\u003ePark Y S, Hong H P, Ryu S R, et al. Effects of textured food masticatory performance in older people with different dental conditions. BMC Geriatr 2022;22(1):384.\u003c/li\u003e\n\u003cli\u003eYang C, Li X, An R, et al. Oral Frailty Assessment Tool: Instrument Development and Cross-Sectional Validation Study. J Oral Rehabil 2025;52(10):1608-1618.\u003c/li\u003e\n\u003cli\u003eLee J Y, Chun A Y, Bae Y, et al. Oral Function as a Predictor of Skeletal Muscle Mass Decline in Community-Dwelling Older Adults. J Oral Rehabil 2025;52(12):2259-2268.\u003c/li\u003e\n\u003cli\u003eLexomboon D, Kumar A, Freyland S, et al. Is poor chewing ability a risk factor for malnutrition? A six-year longitudinal study of older adults in Sweden. J Nutr Health Aging 2025;29(6):100554.\u003c/li\u003e\n\u003cli\u003eStrączek A, Szałkowska J, Sutkowska P, et al. Impact of nutrition on the condition of the oral mucosa and periodontium: A narrative review. Dental and medical problems 2023;60(4):697-707.\u003c/li\u003e\n\u003cli\u003eJu Y J, Lee J E, Lee S Y. Associations between Chewing Difficulty, Subjective Cognitive Decline, and Related Functional Difficulties among Older People without Dementia: Focus on Body Mass Index. J Nutr Health Aging 2021;25(3):347-355.\u003c/li\u003e\n\u003cli\u003eSun Y, Liu H, Li X, et al. Association of social frailty, sarcopenia, and oral frailty with depressive symptoms in Chinese older adults: a cross-sectional study. BMC Public Health 2025;25(1):464.\u003c/li\u003e\n\u003cli\u003eKawamura K, Maeda K, Miyahara S, et al. Oral hypofunction and social aspects in older adults visiting frailty outpatient clinic. J Oral Rehabil 2024;51(10):2150-2157.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTable 1\u0026nbsp;Baseline Characteristics of MCI Patients According to Oral Frailty Status\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"709\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eCharacteristics\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTotal\u003c/p\u003e\n \u003cp\u003e(456)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNon-Oral Frailty(119)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eOral Frailty\u003c/p\u003e\n \u003cp\u003e(337)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cem\u003ex\u003c/em\u003e\u003csup\u003e2\u003c/sup\u003e/\u003cem\u003eZ\u0026nbsp;\u003c/em\u003evalue\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eP value\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSex\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.026\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.871\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003emale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e250(54.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e66(55.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e184(54.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003efemale\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e206(45.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e53(44.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e153(45.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAge,year,median(IQR)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e68.00(63.00,72.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e66.00(62.00,70.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e68.00(64.00,72.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-3.612\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAge group\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e9.012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.011\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e60~69 year\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e275(60.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e85(71.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e190(56.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e70~79 year\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e156(34.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e31(26.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e125(37.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026ge;80 year\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e25(5.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3(2.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e22(6.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eBMI,kg/m\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.303\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.316\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e18.5~23.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e137(30.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e37(31.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e100(29.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e24.0~27.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e182(40.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e41(34.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e141(41.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026ge;28\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e137(30.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e41(34.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e96(28.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMarriage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e4.176\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.124\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMarried\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e422(92.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e115(96.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e307(91.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eDivorced\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e14(3.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1(0.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e13(3.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWidowed\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e20(4.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3(2.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e17(5.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eEducation Level\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e9.553\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.049\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eElementary school or lower\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e123(27.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e23(19.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e100(29.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eJunior high school\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e205(45.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e55(46.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e150(44.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eHigh school\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e86(18.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e32(26.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e54(16.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eAssociate\u0026apos;s degree\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e27(5.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e6(5.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e21(6.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eBachelor\u0026apos;s degree or higher\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e15(3.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e3(2.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e12(3.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eIncome\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e4.639\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.200\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e＜1000 RMB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e65(14.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e11(9.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e54(16.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1000~2999 RMB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e199(43.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e50(42.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e149(44.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3000~4999 RMB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e141(30.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e43(36.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e98(29.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e＞5000 RMB\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e51(11.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e15(12.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e36(10.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMedical insurance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e5.222\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.265\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eResidents\u0026apos; Medical\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e304(66.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e84(70.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e220(65.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNew Rural Cooperative\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e116(25.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e26(21.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e90(26.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSubsistence Allowance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e15(3.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e5(4.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e10(3.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCommercial Insurance\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e7(1.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e3(2.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e4(1.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSelf-pay\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e14(3.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1(0.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e13(3.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eResidential Setting\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e4.972\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.026\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eTowns\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e279(61.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e83(69.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e196(58.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eRural\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e177(38.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e36(30.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e141(41.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLiving Arrangement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e9.858\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.020\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWith spouse and children\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e47(10.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e8(6.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e39(11.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWith spouse\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e317(69.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e96(80.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e221(65.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWith children\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e51(11.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e7(5.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e44(13.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eLive alone\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e41(9.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e8(6.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e33(9.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSmoking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e16.224\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e156(34.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e25(21.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e131(38.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e274(60.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e90(75.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e184(54.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eQuit smoking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e26(5.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e4(3.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e22(6.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eDrinking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3.135\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.209\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e137(30.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e32(26.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e105(31.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e292(64.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e83(69.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e209(62.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eQuit drinking\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e27(5.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e4(3.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e23(6.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eComorbidity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e19.300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0 type\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e87(19.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e37(31.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e50(14.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1~3 type\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e332(72.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e79(66.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e253(75.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e＞3 type\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e37(8.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e3(2.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e34(10.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eSarcopenia,median(IQR)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.00(0.00,3.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1.00(0.00,1.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2.00(0.00,4.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-5.326\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e18.659\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e113(24.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e12(10.1)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e101(30.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e343(75.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e107(89.9)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e236(70.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMalnutrition,median(IQR)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e10.00(8.00,11.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e11.00(10.00,11.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e9.00(8.00,11.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-5.514\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e40.905\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e263(57.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e39(32.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e224(66.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e193(42.3)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e80(67.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e113(33.5)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eChewing Difficulties\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e72.311\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eYes\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e248(54.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e25(21.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e223(66.2)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eNo\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e208(45.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e94(79.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e114(33.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eDenture\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e218.451\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNone\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e181(39.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e115(96.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e66(19.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e1~5 dentures\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e236(51.8)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2(1.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e234(69.4)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e＞5 dentures\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e39(8.6)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e2(1.7)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e37(11.0)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eDepression,median(IQR)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e9.00(7.00,10.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e9.00(8.00,10.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e8.00(7.00,9.00)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e-4.333\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eTable 2 The parameters of the three types of machine learning prediction models\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"678\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eModel\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eAUC(95% CI)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eSensitivity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eSpecificity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eAccuracy\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eF1 score\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eBinary Logistic Regression\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.917(0.855,0.979)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.911\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.964\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.957\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.928\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eDecision Tree\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.907(0.852,0.961)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.861\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.889\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.956\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.906\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eXGBoost\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.917(0.857,0.976)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.821\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.944\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.976\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003e0.892\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eTable 3\u0026nbsp;The logistic regression analysis of oral frailty\u003c/p\u003e\n\u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"690\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 214px;\"\u003e\n \u003cp\u003eVariable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69px;\"\u003e\n \u003cp\u003e\u003cem\u003e\u0026beta;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003eSE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e\u003cem\u003ez\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003eP\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 81px;\"\u003e\n \u003cp\u003eOR\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 124px;\"\u003e\n \u003cp\u003e95% CI\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 214px;\"\u003e\n \u003cp\u003eLiving Arrangement\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69px;\"\u003e\n \u003cp\u003e-3.057\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e1.303\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e-2.346\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e0.019\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 81px;\"\u003e\n \u003cp\u003e0.047\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 124px;\"\u003e\n \u003cp\u003e0.003~0.545\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 214px;\"\u003e\n \u003cp\u003eMalnutrition\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69px;\"\u003e\n \u003cp\u003e2.089\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e0.550\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e3.795\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 81px;\"\u003e\n \u003cp\u003e8.076\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 124px;\"\u003e\n \u003cp\u003e2.909~25.738\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 214px;\"\u003e\n \u003cp\u003eChewing Difficulty\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69px;\"\u003e\n \u003cp\u003e2.489\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e0.532\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e4.681\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 81px;\"\u003e\n \u003cp\u003e12.055\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 124px;\"\u003e\n \u003cp\u003e4.483~36.770\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 214px;\"\u003e\n \u003cp\u003eDenture\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69px;\"\u003e\n \u003cp\u003e5.082\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e0.770\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e6.596\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e\u0026lt; 0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 81px;\"\u003e\n \u003cp\u003e161.016\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 124px;\"\u003e\n \u003cp\u003e42.603~920.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\" style=\"width: 214px;\"\u003e\n \u003cp\u003eDepression\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 69px;\"\u003e\n \u003cp\u003e-0.407\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 59px;\"\u003e\n \u003cp\u003e0.155\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 77px;\"\u003e\n \u003cp\u003e-2.631\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 67px;\"\u003e\n \u003cp\u003e0.008\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 81px;\"\u003e\n \u003cp\u003e0.665\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\" style=\"width: 124px;\"\u003e\n \u003cp\u003e0.484~0.893\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-geriatrics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bgtc","sideBox":"Learn more about [BMC Geriatrics](http://bmcgeriatr.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bgtc/default.aspx","title":"BMC Geriatrics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Oral frailty, oral health, mild cognitive impairment, cognitive impairment, older adults","lastPublishedDoi":"10.21203/rs.3.rs-8962009/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8962009/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eObjectives\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOral frailty is a multidimensional geriatric syndrome associated with adverse health outcomes. Older adults with mild cognitive impairment (MCI) may be particularly vulnerable; however, prediction tools tailored to this population remain limited. This study aimed to develop and internally validate a risk prediction model for oral frailty in older adults with MCI.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDesign\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCross-sectional study for model development.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSetting and Participants\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA total of 456 older adults diagnosed with MCI were recruited from a geriatric outpatient clinic between October 2025 and January 2026. Participants were randomly divided into a training cohort (70%) and an independent validation cohort (30%).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOral frailty was assessed using the Oral Frailty Index-8. Candidate predictors included demographic, clinical, nutritional, and psychosocial variables. Feature selection was performed using random forest and least absolute shrinkage and selection operator regression. Logistic regression, decision tree, and extreme gradient boosting models were developed. Model discrimination, calibration, and clinical utility were evaluated in the validation cohort.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe prevalence of oral frailty was 73.9%. In the validation cohort, the logistic regression model demonstrated good discrimination (AUC 0.917, 95% CI 0.855–0.979) and satisfactory calibration. Decision curve analysis indicated favorable net clinical benefit across a range of threshold probabilities. Independent predictors included denture burden (OR 161.016, 95% CI 42.603–920.001), chewing difficulty (OR 12.055, 95% CI 4.483–36.770), malnutrition (OR 8.076, 95% CI 2.909–25.738), depressive symptoms (OR 0.665, 95% CI 0.484–0.893), and living arrangement (OR 0.047, 95% CI 0.003–0.545).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions and Implications\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOral frailty among older adults with MCI is strongly associated with structural oral impairment, functional decline, nutritional vulnerability, and psychosocial factors. The proposed model demonstrates good predictive performance and may facilitate early identification of high-risk individuals in clinical settings. Integrated oral–cognitive assessment strategies may improve comprehensive geriatric care for this vulnerable population.\u003c/p\u003e","manuscriptTitle":"Development and Validation of a Risk Prediction Model for Oral Frailty in Older Adults With Mild Cognitive Impairment: A Cross-Sectional Study","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-28 16:26:02","doi":"10.21203/rs.3.rs-8962009/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewersInvited","content":"","date":"2026-03-25T11:04:23+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-23T08:53:38+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-02-27T09:29:24+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-02-27T02:44:06+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Geriatrics","date":"2026-02-26T06:04:34+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-geriatrics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"bgtc","sideBox":"Learn more about [BMC Geriatrics](http://bmcgeriatr.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/bgtc/default.aspx","title":"BMC Geriatrics","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a8387b6f-8fb5-40a4-a6c7-7af23bd8f2ec","owner":[],"postedDate":"March 28th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-03-28T16:26:02+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-28 16:26:02","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8962009","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8962009","identity":"rs-8962009","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}