Prediction Model for Frailty in Middle-Aged and Older Adults with Cardiovascular Disease

Introduction

Cardiovascular disease (CVD) is a major cause of death in middle-aged and elderly populations, posing a significant burden on global public health.^1,2 According to the World Health Organization (WHO), approximately 17.9 million people died from cardiovascular disease worldwide in 2019, accounting for about 32% of all global deaths.³ With advancing age, the prevalence of CVD increases substantially, and patients with coexisting geriatric syndromes are more likely to experience adverse cardiovascular events and poorer prognoses.

Alongside global population aging, frailty has emerged as a common and clinically significant geriatric syndrome and is increasingly recognized as a key determinant of health outcomes among middle-aged and older adults.⁴ Frailty is an age-related multidimensional clinical syndrome characterized by a decline in physiological reserve across multiple organ systems, resulting in increased vulnerability to stressors and reduced resilience.⁵ The frailty phenotype proposed by Fried et al defines frailty based on five core domains: unintentional weight loss, reduced grip strength, exhaustion, slowed walking speed, and low physical activity.⁶ This definition has been widely adopted in both clinical and epidemiological studies. Accumulating evidence indicates that frailty is strongly associated with a range of adverse health outcomes, including prolonged hospitalization, increased readmission rates, falls, disability, and higher all-cause mortality, making it an important indicator for assessing health status and prognosis in older adults.^7–10 In recent years, the clinical significance of frailty in patients with CVD has received increasing attention. Compared with individuals without CVD, patients with CVD exhibit a significantly higher prevalence of frailty, which has been identified as an independent predictor of mortality and adverse outcomes, with mortality risk increasing by approximately twofold.^11–14 CVD and frailty share substantial pathophysiological overlap, including chronic low-grade inflammation, mitochondrial dysfunction, oxidative stress, metabolic dysregulation, and aging-related biological changes. These shared mechanisms support the concept of frailty as an important functional phenotype reflecting disease progression and prognosis in CVD.¹¹ Moreover, many of these biological processes can be indirectly captured through clinically measurable indicators, providing a theoretical basis for developing prediction models based on routine clinical variables.

The prevalence of frailty among patients with CVD varies across regions. A European population-based study reported a frailty prevalence of approximately 43% among patients with CVD, whereas a study of hospitalized Chinese patients with CVD found a prevalence of about 28.0%.^15,16 These findings suggest that frailty is common in the CVD population and may be influenced by regional, ethnic, and healthcare system differences, highlighting the need for population-specific investigations.

Despite the recognized prognostic importance of frailty in CVD, frailty assessment in routine cardiovascular practice still primarily relies on diagnostic tools such as comprehensive geriatric assessment (CGA), which are time-consuming and difficult to implement in outpatient screening or community follow-up settings.¹⁷ As CVD management shifts from event-driven treatment toward risk prediction and early intervention, there is an urgent need for simple and scalable tools to predict frailty risk. Existing frailty prediction models are mostly developed in general older populations or depend on complex assessments and laboratory measurements, limiting their feasibility and applicability in routine clinical practice for CVD patients. Therefore, a practical prediction model based on easily obtainable clinical indicators specifically tailored for middle-aged and older adults with CVD remains lacking.

Accordingly, this study utilized data from the China Health and Retirement Longitudinal Study (CHARLS) to develop and validate a nomogram model for predicting frailty risk among middle-aged and older patients with CVD. The proposed model aims to provide a simple and visual risk assessment tool for outpatient screening, long-term follow-up monitoring, and individualized intervention decision-making, thereby facilitating early identification of high-risk individuals and improving long-term health outcomes.

Materials and Methods

Study Design and Participants

This study used data from the 2015 CHARLS and was designed as a cross-sectional study. CHARLS is a nationwide prospective cohort study that aims to collect high-quality data on middle-aged and elderly people aged 45 and above and their families, to analyze the characteristics of population aging and promote related interdisciplinary research. CHARLS research protocol was approved by the Peking University Biomedical Ethics Committee (IRB00001052-11015), and all participants signed an informed consent form before enrollment.

Inclusion criteria: (1) age ≥ 45 years; (2) diagnosis of CVD.

Exclusion criteria: (1) age < 45 years; (2) absence of frailty data; (3) missing covariate data. Following the screening process, 1184 participants were enrolled in the study. The inclusion and exclusion process of the study participants is shown in Figure 1.

Figure 1 Flowchart.

Data Collection

Definition of Frailty

The frailty phenotype is a widely recognized and valid diagnostic criterion for assessing frailty. It defines frailty as a clinical syndrome characterized by five indicators of physical frailty: unintentional weight loss, weakness or low grip strength, exhaustion, slow walking speed, and low physical activity.⁶ In the present study, a modified operational definition based on the Fried frailty phenotype was adopted. Considering the variables available in the CHARLS database, appropriate adaptations were made to the original criteria. Previous studies have demonstrated that such modified assessments show good validity in identifying frailty.¹⁸ Frailty was evaluated across five domains: weakness, exhaustion, slowness, low physical activity, and weight loss. Each component was coded as “1” or “0”, yielding a total score ranging from 0 to 5. Participants with a total score ≥3 were classified as frail.^6,19 The specific evaluation criteria are as follows:

Weakness

Participants were classified as having weakness if they self-reported difficulty lifting or carrying objects weighing more than 5 kg.¹⁹

Fatigue

Fatigue was assessed using the short version of the Center for Epidemiologic Studies Depression Scale (CES-D). Participants were considered to have fatigue if they responded “most or all of the time” or “occasionally or a moderate amount of time” to either of the following statements: “I felt that everything I did was an effort” or “I could not get going”.⁶ The Chinese version of the CES-D has been widely used among middle-aged and older Chinese populations and has demonstrated good reliability and validity.²⁰

Slowness

Slowness was defined based on the criterion of reduced walking speed in the Fried frailty phenotype. As objective gait speed measurements were not available in the database, self-reported items related to physical functional limitations from the CHARLS questionnaire were used as proxy indicators. Participants who reported difficulty walking 100 meters or climbing several flights of stairs were classified as having slowness, consistent with methods used in previous studies.²¹

Insufficient Physical Activity

Insufficient physical activity was assessed using relevant items from the CHARLS questionnaire regarding the frequency and duration of daily physical activities. Participants were classified as having insufficient physical activity if they reported not engaging in moderate-intensity physical activity or at least 10 minutes of walking per week. Although this variable differs from the original criterion proposed by Fried et al, similar indicators have been used to assess frailty in previous studies, and this operational definition has been applied in prior research based on the CHARLS database.²¹

Weight Loss

Weight loss was defined according to the criterion of unintentional weight loss in the Fried frailty phenotype. Participants were classified as having weight loss if they reported unintentional weight loss of ≥5 kg during the past year or had a current body mass index (BMI)≤18.5 kg/m2.²²

Definition of Cardiovascular Disease Events

The study outcome was the presence of CVD, including heart disease and stroke, consistent with previous studies.^23,24 In CHARLS, CVD was assessed based on participants’ self-reported physician diagnoses. Participants were asked whether a doctor had ever informed them that they had been diagnosed with heart attack, coronary heart disease, angina, congestive heart failure, or other heart problems, or whether they had been diagnosed with stroke. Participants who provided affirmative responses to any of these questions were classified as having CVD.

Independent Variable

This study analyzed sociodemographic characteristics, including age, sex, education level, residence (urban or rural), and marital status. Lifestyle behaviors included smoking history, alcohol consumption history, number of cigarettes smoked per day, drinking frequency, participation in social activities, sleep quality, and nighttime sleep duration. Health conditions factors included waist circumference, vision, hearing, activities of daily living (ADL) score, pain, grip strength, chronic diseases, and cognitive function. Cognitive function was assessed using the standardized CHARLS cognitive assessment module, which includes four domains: episodic memory (immediate and delayed recall), time orientation, numerical ability, and figure drawing. The total score ranged from 0 to 21, with higher scores indicating better cognitive function. Psychological factors that were considered in this study encompassed both depression and overall life satisfaction. The 10-item CES-D was used to assess depression. Participants with a score greater than 10 were considered to have depressive symptoms.²⁵

Statistical Analysis

This study was conducted using data from the 2015 wave of the CHARLS. Missing data were handled using a complete-case analysis approach. Participants with missing information on the frailty outcome or key covariates were excluded prior to variable selection and model construction. Continuous variables were presented as mean ± standard deviation (SD) if normally distributed, or as median and interquartile range (IQR) if non-normally distributed. Categorical variables were expressed as frequencies and percentages. Group comparisons were performed using the χ² test for categorical variables and the independent-samples t test or Mann–Whitney U-test for continuous variables, as appropriate based on data distribution.²⁶

To develop the prediction model, the total sample was randomly divided into a training set and a validation set at a ratio of 7:3. The training set was used for variable selection, model development, and parameter estimation, while the validation set was used for independent model validation. In the training set, the least absolute shrinkage and selection operator (LASSO) regression was first applied to identify potential predictors associated with frailty among patients with CVD. To eliminate the influence of different measurement scales on variable selection, all continuous variables were standardized using Z-score transformation prior to LASSO regression. The optimal penalty parameter (λ) was determined using ten-fold cross-validation, and variables with nonzero regression coefficients were selected as candidate predictors. Subsequently, the selected variables were entered into a multivariable logistic regression model. Variables with P < 0.05 were considered independent predictors of frailty and were used to construct a nomogram prediction model based on the regression coefficients. Model discrimination was evaluated using the area under the receiver operating characteristic (ROC) curve (AUC), with higher AUC values indicating better predictive performance. The 95% confidence intervals (CIs) for the AUC were calculated using a nonparametric bootstrap method with 1000 resamples. Model calibration was assessed using calibration curves and the Hosmer–Lemeshow goodness-of-fit test to evaluate agreement between predicted probabilities and observed outcomes. Calibration curves were generated using bootstrap resampling. Clinical utility was evaluated using decision curve analysis (DCA), which estimates the net benefit of the model across different threshold probabilities to assess its potential value in clinical decision-making. All statistical analyses were performed using R software (version 4.4.1). All tests were two-sided, and P < 0.05 was considered statistically significant.

Result

Participant Characteristics

A total of 1184 patients with CVD were included in the final analysis. Among the included CVD patients, 148 (12.5%) had frailty symptoms. Significant differences were identified regarding age, marital status, social activity involvement, life satisfaction, and sleep duration between frail and non-frail patients. The demographic and clinical characteristics of the participants are meticulously summarized in Table 1. All CVD patients were randomly divided into a training set (n = 829) and a validation set (n = 355) at a ratio of 7:3. The detailed baseline characteristics of the two groups are provided in Table S1. No significant differences were observed between the two groups in baseline characteristics except for waist circumference, hypertension, and hearing status (P > 0.05).

Table 1 Baseline Characteristics of Participants

Construction of Predictive Models

LASSO regression was applied to select variables, and those with nonzero regression coefficients were identified as potential predictors of frailty among patients with CVD (Figure 2A and B). The selected variables were subsequently entered into a multivariable logistic regression model using the “rms” package in RStudio for further analysis. The results showed that cognitive function, ADL score, sleep duration, grip strength, age, waist circumference, and depression were significantly associated with frailty in patients with CVD (Table 2).

Table 2 Multivariable Logistic Regression Analysis of Predictors of Frailty

Figure 2 Variable selection using LASSO regression. (A) Coefficient profiles of the LASSO model. The optimal tuning parameter λ was selected by plotting log(λ) on the x-axis and the corresponding regression coefficients on the y-axis, with each colored curve representing a different variable. (B) The optimal λ was determined using 10-fold cross-validation based on the minimum criterion. The plot shows the binomial deviance (partial likelihood deviance) against log(λ), with a vertical dashed line indicating the λ value selected according to the 1-standard-error (1-SE) rule.

Development of Predictive Models

The variance inflation factor (VIF) test confirmed that all values were under 4, demonstrating that multicollinearity was not present among the variables. The selected independent predictors were cognitive function, ADL score, sleep duration, grip strength, age, waist circumference, and depression. Based on these significant predictors, a nomogram was constructed to quantitatively estimate the risk of frailty among CVD patients (Figure 3).

Figure 3 Nomogram.

Validation of Predictive Models

The predictive model’s discriminative performance was assessed by computing the area under the AUC for the training and validation datasets (Figure 4A and B). The AUC value of the model was 0.851 (95% CI: 0.814–0.888) in the training set and 0.861 (95% CI: 0.804–0.917) in the validation set. Both AUC values exceeded 0.8, indicating that the nomogram demonstrated strong discrimination and predictive performance, effectively distinguishing between frail and non-frail patients with CVD.

Figure 4 (A) Nomogram ROC curve based on the training dataset. (B) Nomogram ROC curve based on the validation dataset.

Calibration

Model calibration was evaluated using calibration curves and the Hosmer–Lemeshow goodness-of-fit test, with P > 0.05 indicating adequate model fit. The results showed no significant deviation between predicted and observed outcomes in either the training set (χ² = 6.67, df = 8, P = 0.5724) or the validation set (χ² = 9.16, df = 8, P = 0.3294), suggesting good overall model fit. The calibration curves demonstrated good agreement between predicted probabilities and observed event rates within the moderate-risk range. However, some deviations between predicted and observed probabilities were observed in the extreme low- and high-risk ranges. In both the training and validation sets, the bias-corrected and apparent calibration curves generally followed the ideal diagonal line, although deviations were evident at the extremes of predicted risk (Figure 5A and B).

Figure 5 (A) Calibration plot for the training dataset. (B) Calibration plot for the validation dataset.

Clinical Effectiveness Evaluation

DCA was performed to evaluate the clinical utility of the prediction model. The results showed that, within threshold probability ranges of 0.1–0.5 in the training set and 0.1–0.8 in the validation set, the model yielded a higher net benefit than both the “treat-all” and “treat-none” strategies. Across these threshold ranges, the model curve consistently remained above the two extreme strategies, indicating that applying the model for risk assessment could provide greater clinical net benefit within these intervals (Figure 6A and B).

Figure 6 (A) DCA curves for the training dataset. (B) DCA curves for the validation dataset.

Discussion

Frailty can negatively affect various body functions, thereby increasing the risk of major adverse cardiovascular events.²⁷ In this study, the incidence of frailty among patients with CVD was 12.5%, which aligns with previously reported ranges of 10.8%–48.5%.²⁸ Given that frailty is an indicator of adverse prognosis in individuals with CVD, early and effective recognition of high-risk patients is important to improve patient outcomes.

This study shows that sleep duration independently predicts frailty in individuals with CVD. The results indicated that patients with shorter nighttime sleep were more likely to develop frailty, consistent with previous studies.^29,30 Chronic short sleep duration or insomnia (sleep duration <6 hours) increases susceptibility to frailty by increasing feelings of fatigue and reducing physical performance in daily activities.^31,32 In addition, our study found that prolonged sleep (sleep duration≥ 9 hours) was also associated with an increased risk of frailty, suggesting a potential U-shaped relationship between sleep duration and frailty.³²

Several mechanisms may explain this association. First, chronic inflammation may play a central role. Both insufficient and excessive sleep can elevate pro-inflammatory cytokines, including interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), and C-reactive protein (CRP).³³ Given that patients with CVD already exhibit chronic inflammatory activation, abnormal sleep patterns may further increase inflammatory burden, promoting muscle protein catabolism, inhibiting synthesis, and accelerating functional decline—key processes in frailty development.^34,35 Second, neuroendocrine dysregulation may contribute to frailty risk. Sleep deprivation disrupts hypothalamic–pituitary–adrenal (HPA) axis function and increases cortisol levels, which are associated with both frailty and CVD risk.³⁶ Increased cortisol has been recognized as a risk factor for both frailty and CVD.^36,37 Reduced sleep duration may also suppress anabolic hormones such as growth hormone, insulin-like growth factor-1 (IGF-1), and testosterone, thereby impairing protein synthesis and promoting sarcopenia.³⁸ Third, sleep abnormalities may impair mitochondrial biogenesis and oxidative phosphorylation, reducing skeletal muscle energy production efficiency.³⁹ In patients with CVD, existing ischemia and oxidative stress may exacerbate mitochondrial dysfunction, accelerating declines in muscle endurance and physical performance.⁴⁰ Behavioral mechanisms may further explain these findings. Prolonged sleep duration is often accompanied by reduced daytime activity and physical exercise, increasing risks of falls, fractures, and sarcopenia—established contributors to frailty.^32,41 Thus, long sleep duration may reflect underlying disease burden or functional decline rather than a protective behavior. Abnormal sleep duration may therefore act both as a risk factor and a clinical marker of deteriorating health status. Approximately 35% of middle-aged and older adults in China experience sleep disturbances, yet these changes are often overlooked or considered part of normal aging.⁴¹ Recognizing the significant association between sleep disturbances and frailty can help healthcare professionals identify high-risk populations early and develop targeted prevention and intervention strategies to improve sleep quality and reduce the risk of frailty in middle-aged and older patients with CVD.

This study also demonstrated a significant association between low ADL scores and frailty in patients with CVD. Patients with lower ADL scores had a higher risk of frailty, consistent with previous studies, further highlighting ADL as an important predictor of frailty.^42,43 ADL represents an individual’s capacity to independently manage routine daily activities. Patients with lower ADL scores have diminished self-care abilities, which can negatively impact their daily nutrition and potentially result in long-term malnutrition. Moreover, impaired ADL restricts patients’ engagement in daily activities and physical exercise, potentially leading to decreased muscle mass, reduced muscle function, and lower physical performance. This increases susceptibility to sarcopenia, fractures, and falls, further heightening the risk of frailty.^43,44 Therefore, routine assessment of ADL in patients with CVD could assist healthcare professionals in early identification and stratification of frailty risk, supporting the implementation of effective strategies to reduce frailty and its related adverse outcomes.

This study demonstrated that grip strength and waist circumference were independent predictors of frailty among patients with CVD. Patients with lower grip strength were more likely to develop frailty. Grip strength is widely recognized as an important indicator of muscle strength and overall physical function. Reduced grip strength not only reflects declines in skeletal muscle mass and strength but is also associated with decreased bone density and increased risks of falls and fractures.^45,46 In addition, grip strength serves as a proxy for nutritional status, and lower values often indicate malnutrition, providing important pathophysiological support for our findings.⁴⁷ Similarly, we found that smaller waist circumference was associated with a higher risk of frailty in patients with CVD, consistent with previous studies.²¹ Among middle-aged and older individuals with CVD, reduced waist circumference may reflect wasting, loss of adipose and muscle tissue, and underlying malnutrition rather than a simple reduction in adiposity. As a chronic systemic disease, CVD can induce persistent activation of inflammatory pathways and neuroendocrine stress responses, leading to an imbalance between protein synthesis and catabolism and promoting the loss of fat and skeletal muscle mass.⁴⁸ This chronic catabolic state may result in weight loss, reduced waist circumference, and decreased muscle mass, thereby increasing the risks of sarcopenia and frailty.^48,49 Importantly, reduced waist circumference in this population may represent diminished energy reserves and muscle depletion rather than merely reduced fat mass. Adipose tissue functions not only as an energy storage organ but also as an endocrine organ that secretes cytokines and hormones essential for metabolic homeostasis.⁵⁰ Excessive depletion of fat reserves may impair the body’s ability to withstand acute stress and chronic disease burden, thereby increasing vulnerability to frailty.⁵¹

Considering the potential confounding effect of BMI on the relationship between waist circumference and frailty, BMI-related variables were included as candidate predictors during variable selection and multivariable logistic regression analyses. After LASSO selection and multicollinearity assessment, waist circumference remained an independent predictor after adjustment for other covariates, suggesting that its association with frailty was not fully explained by BMI. Compared with BMI, which reflects overall body weight, waist circumference may better capture abdominal energy reserves and muscle status and may more sensitively reflect nutritional depletion and metabolic abnormalities in older adults with chronic diseases. Overall, reduced grip strength and smaller waist circumference may jointly indicate sarcopenia and chronic catabolic status in patients with CVD, representing important pathological foundations of frailty. Early nutritional support, resistance training, and individualized exercise interventions may therefore help reduce frailty risk among individuals with impaired muscle function or nutritional deficiency.

This study revealed a significant association between frailty and diminished cognitive function, consistent with previous research.⁵² The link between frailty and cognitive decline is based on the fact that both share many pathophysiological mechanisms, such as inflammation, oxidative stress response, neuroendocrine dysfunction, mitochondrial disorders.¹³ These pathological mechanisms also play crucial roles in the onset and progression of CVD. Studies indicate that cognitive dysfunction is more common in patients with CVD than in those without, and among individuals with frailty, its prevalence is higher than in those without frailty.¹³ Moreover, CVD acts as an important risk factor for cognitive decline, accelerating the deterioration of cognitive function. Cognitive impairment, in turn, may reduce patients’ self-management ability and treatment adherence, leading to disease progression and an increased risk of frailty. Therefore, attention should be given to the cognitive function of CVD patients, and cognitive rehabilitation training should be provided as early as possible to patients with cognitive decline, so as to delay the progression of the disease and help improve frailty.

The results suggest that age is an important predictor of frailty among patients with CVD, with frailty risk increasing progressively with advancing age. Aging is associated with a progressive decline in physiological function across multiple organ systems, leading to increased vulnerability. The underlying mechanisms include telomere dysfunction, chronic inflammation, dysregulation of the adaptive immune system, reduced muscle mass and function, and cognitive impairment.⁵³ Furthermore, studies have shown that reduced physical activity with aging contributes to decreased muscle strength and bone mineral density, increasing the risk of osteoporosis, falls, and fractures, which in turn promote frailty. In addition, studies have shown that aging is associated with decreased bone mineral density and muscle strength, which increase the risk of osteoporosis, falls, and fractures, further contributing to the development of frailty.^54,55 Therefore, early identification and targeted intervention for older adults at high risk of CVD are essential. Implementing appropriate preventive and rehabilitative strategies help slow or even reverse the progression of frailty and reduce the risk of adverse outcomes.

This study revealed a link between depression and frailty, consistent with prior findings that depressive symptoms heighten frailty risk.⁵⁶ Several mechanisms may underlie this association. Depression often leads to reduced social engagement, sleep disturbances, weight loss, and loss of appetite, all of which negatively affect physical health and increase vulnerability to frailty.⁵⁶ In addition, depression and frailty may share common biological pathways, including chronic inflammation, oxidative stress, and mitochondrial dysfunction,⁵⁷ which further contribute to the development of frailty. Moreover, depression has been shown to increase the risk of CVD by 80–90% and to negatively influence disease progression and prognosis.⁵⁸ Therefore, it is necessary to screen for depression among individuals at high risk of CVD and to provide timely psychological support and counseling for patients with depressive symptoms, thereby reducing the risk of adverse outcomes.

Nomograms are widely used clinical prediction tools that quantify the relative contribution of risk factors through a scoring system and enable individualized risk estimation. The nomogram developed in this study, based on seven independent predictors, demonstrated good discrimination and calibration in internal validation. The area under the AUC exceeded 0.80, indicating strong discriminative ability and accurate differentiation between frail and non-frail individuals. DCA further showed positive net benefits across a wide range of threshold probabilities, suggesting that the model may outperform both “treat-all” and “treat-none” strategies in clinical decision-making and therefore has potential clinical applicability. Notably, the seven predictors are unlikely to act independently but may jointly influence frailty development through interconnected pathophysiological pathways. For example, chronic inflammation and metabolic dysregulation may simultaneously affect sleep duration, grip strength, waist circumference, and cognitive function. Depression may indirectly promote frailty progression by impairing sleep and reducing activity levels, while declines in ADL may represent both manifestations and accelerators of functional deterioration. Although multivariable regression was used to control for confounding factors, potential synergistic or additive interactions among predictors may still exist. Future studies using pathway analysis or structural equation modeling may help further clarify these complex interaction mechanisms.

The prediction model is applicable to middle-aged and older adults aged ≥ 45 years with confirmed CVD, particularly in community follow-up settings, primary healthcare institutions, and cardiovascular outpatient screening. All variables included in the model are readily obtainable clinical indicators that do not require complex laboratory testing, supporting its practicality and feasibility in routine clinical practice. Early identification of high-risk individuals may facilitate timely implementation of comprehensive management strategies, including nutritional support, exercise rehabilitation, and psychological interventions, thereby potentially reducing the risk of frailty.

Several limitations of this study should be acknowledged. First, the diagnosis of CVD was based on self-reported physician diagnoses, which may be subject to recall bias. Second, some potential confounding factors were not available in the database, including dietary patterns, medication use, and inflammatory biomarkers, which may have influenced the accuracy of the results. Third, as this was a cross-sectional analysis, only associations could be identified, and causal relationships cannot be inferred. Fourth, external validation of the model has not yet been performed, and its generalizability requires further evaluation in diverse populations. Future multicenter prospective studies are needed to conduct external and dynamic validation to improve model robustness and clinical applicability.

Conclusion

Based on data from the CHARLS, this study developed and validated a nomogram model for predicting frailty risk among patients with CVD. Seven independent predictors were identified, including sleep duration, ADL, waist circumference, cognitive function, grip strength, age, and depression. The model demonstrated good discrimination, calibration, and clinical net benefit, enabling individualized quantitative estimation of frailty risk. These factors may jointly contribute to frailty development through shared mechanisms involving inflammation, neuroendocrine dysregulation, and metabolic dysfunction, suggesting common pathological pathways between CVD and frailty. In the context of global population aging, this model provides a feasible tool for early screening in both clinical and community settings. Nevertheless, multicenter prospective studies with external validation are required to further improve the model’s generalizability.