Abstract
Objectives
To develop and independently externally validate a novel, simple, and validated tool to assess the risk of cognitive frailty in older adult patients with heart failure based on these factors.
Design
A cross-sectional study.
Data sources
A novel nomogram risk prediction model was developed by recruiting older adult patients with heart failure for data collection from October 2022 to August 2023 at a university general hospital in China. Independent external validation of the developed model was performed on patients collected from March 2023 to June 2023 at a university general hospital at the same level.
Methods
Univariate and multivariate logistic regression analyses were performed for variables that may influence the prevalence of cognitive frailty. The fit and predictive performance of the nomogram risk prediction model was evaluated based on the area under the characteristic curve of the subjects and decision curve analysis.
Results
The overall prevalence of cognitive frailty in older adult patients with heart failure was 44.5%. The results showed that the predictive ability of the model development cohort was 0.807(95%CI = 0.7640.851, P<0.001). And the predictive ability of the validation cohort was 0.770(95%CI = 0.6670.874, P<0.001), indicating that the model has high accuracy.
Conclusions
The results showed satisfactory predictive performance. The use of this tool helps clinical caregivers to quickly and intuitively identify patients at high risk for the development of cognitive frailty on admission, thus providing evidence to support the early development and implementation of personalized strategies.
Keywords: Older adult, Heart failure, Cognitive frailty, Nomogram, Risk prediction
What is already known
Cognitive frailty is highly prevalent in older adult patients with heart failure and has a significant impact on patient quality of life and healthcare resource utilization, yet caregivers’ awareness of it is currently low and inadequately assessed.
Early identification of older adult patients with heart failure who are at risk for developing cognitive frailty can help to implement early intervention and high-quality care.
To date, there are no predictive tools related to the risk of cognitive frailty in older adult patients with heart failure.
What this paper adds
Age, NYHA classification, chewing ability of teeth, nutritional status, loneliness, and sleep status are independent predictors of the development of cognitive frailty in older adult patients with heart failure.
A novel, simple nomogram risk prediction model with good predictive performance was developed and independently externally validated to effectively assess the risk of cognitive frailty in older adult patients with heart failure.
Introduction
Heart failure is now recognized as a global epidemic, afflicting approximately 64 million patients worldwide [1]. According to the latest statistics, the disease is particularly prevalent in adults aged 60 years and older, with prevalence rates climbing to 10 ~ 20% [2]. It is important to note that as the global population ages, the prevalence of heart failure in older adults will continue to rise [3].
Several studies have shown that the prevalence of physical frailty in older adult patients with heart failure is significantly increased, which make the patients less resistant to stress [4]. A meta-analysis showed that the overall prevalence of physical frailty in older adult patients with heart failure was 44.5% [5]. A study showed that patients with comorbid physical frailty showed a significant increase in mortality, and hospitalization costs [6]. Another non-negligible challenge faced by older adult patients with heart failure in recent years is their impaired cognitive function. A study showed that older adult patients with heart failure had a fourfold higher risk of developing mild cognitive impairment than normal older adult, with a prevalence between 13.5%~63.4% [7]. A systematic review showed that mild cognitive impairment significantly increased the risk of death in older adult patients with heart failure, with a 1-month mortality rate of more than 25% [8].
In 2013, Kelaiditi et al. first elucidated the new concept of cognitive frailty (CF), where physical frailty coexists with mild cognitive impairment and excludes various types of neurodegenerative pathologies [9]. Fortunately, however, both conditions are amenable to potentially reversible interventions [10]. Most previous studies have examined physical frailty and mild cognitive impairment in older adult patients with heart failure as two separate entities. However, several recent studies have demonstrated that physical frailty is known to increase the risk of cognitive impairment, whereas impaired cognitive function may increase the risk of physical frailty and have an impact on the trajectory of physical frailty [11, 12].
Older adult patients with heart failure have many symptoms, and when the two coexist, they significantly increase the adverse clinical outcomes of patients. In addition, the evaluation methods for cognitive frailty have not been standardized, and lacking specific biochemical indicators to consider older adult patients with heart failure in the assessment process [13]. NYHA classification and NT-proBNP have been found to be significantly associated with physical frailty and cognitive dysfunction in older adult patients with heart failure, respectively [14, 15]. Therefore, universal tools for the assessment of the risk of developing cognitive frailty in older adult patients with heart failure have become insufficient, and there is an urgent need to construct specialized and more specific assessment tools.
In recent years, determining the probability of an individual’s occurrence of a certain event by integrating multiple predictors and weighting the predictors according to their weights has received attention from scholars, and accurate and efficient results can be obtained through this tool, which is now widely used in the medical field [16]. Although the performance of these models performs well, there is mostly no independent external validation, which makes the results may have some bias. Therefore, to address the limitation that there is no specific assessment tool for cognitive frailty in older adult patients with heart failure, the present study aimed to construct a model for predicting the risk of cognitive frailty in such patients and to conduct independent external validation. To provide clinical caregivers and social workers with a convenient, rapid, and effective tool for assessing the risk of developing cognitive frailty in hospitalized older adult patients with heart failure, in order to identify risk factors and take interventions, and thus to provide a scientific basis for decision-making to improve the adverse clinical outcomes of patients.
Materials and methods
Study design and participants
Data were collected on older adult patients with heart failure from October 2022 to August 2023 at a university-affiliated general hospital to develop a new risk prediction model. Independent external validation of the developed model was performed on patients who met the inclusion criteria collected from March 2023 to June 2023 at a university general hospital at the same level.
Patients were included in this study if they; (a) were diagnosed with heart failure by a clinician according to the Chinese Guidelines for the Diagnosis and Treatment of Heart Failure [17], (b) were ≥ 60 years of age, and (c) were able to cooperate in completing the test. Exclusion criteria were; (a) verbal communication difficulties, (b) coma or in the terminal stage of the disease, and (c) neurodegenerative disease.
Based on the sample size requirements for developing a clinical predictive model, the primary outcome measure in this study was cognitive frailty. For a categorical outcome variable, the required sample size must be estimated according to the number of predictors [18]. With six predictors in this study, the required sample size ranged from 10 to 20 times the number of predictors, equating to 60 to 120. Ultimately, the study enrolled 390 patients in the training set and 80 patients in the independent external validation set, meeting the required sample size.
Data collection
The study used a questionnaire method of data collection in two hospitals according to different times, and the process and content of the collection were strictly in accordance with the same criteria in order to provide a valid external validation of the model. Prior to the formal start of the survey, guidance was provided by qualified assessment professionals, and the researchers received uniform training in order to effectively use the tool to conduct a comprehensive assessment of the study population. If any key variables were found to be missing from the questionnaire, the questionnaire was judged to be invalid and excluded. In organizing the questionnaires, a uniform numbering system was adopted and a two-person data entry technique was used.
Candidate predictor variables and instruments
Basic information included socio-demographic information such as gender and age; Information on biochemical markers such as NYHA classification, NT-proBNP, and left ventricular ejection fraction.
Nutritional status was assessed using the Short-Form Mini Nutritional Assessment. With a score range of 0 to 14, individuals showed better nutritional status as the total score increased [19]; Sleep status was measured using the Asens Insomnia Scale, with individual scores ranging from 0 to 3 and total scores ranging from 0 to 24, with the higher the total score, the poorer the patient’s sleep quality [20]; and Loneliness was assessed using the Loneliness Scale, resulting in a total scale score of 8 to 32, with the higher the score, the stronger the feeling of loneliness [21]; and Depression was measured using the 5-item Geriatric Depression Scale, when the score is greater than or equal to 2, it can be inferred that there are depressive symptoms [22]; Social support is measured by Social Support Rating Scale, the total score ranges from 12 to 66, the higher the overall score, it means that the social support is better [23].
Criteria for rating cognitive frailty in older adult patients with heart failure
The state of physical frailty was determined using the FRAIL Frailty Scale, which is assessed in five dimensions. The composite score for each entry ranged from 0 to 5. If the score obtained is 0, the state is judged to be normal; if the score is between 1 and 2, the body is judged to be in the early stages of frailty; and if the score obtained is more than or equal to 3, the state is judged to be a state of frailty in bodily functions [24, 25].
Cognitive function was assessed using the Montreal Cognitive Assessment Scale (MoCA). The total score ranges from 0 to 30, with a score of 26 or higher indicating “normal” cognitive function. The MoCA total score takes into account years of education, and, as recommended in the MoCA manual, an additional 1 point is added to the test result for those with less than 12 years of education [26].
According to the 2013 International Consensus Panel definition [9], cognitive frailty refers to a condition in which there is a combination of both physical frailty and mild cognitive impairment, but Alzheimer’s disease or other types of dementia are excluded. In this study, dementia and other neurodegenerative diseases were all definitively diagnosed by a team of clinical neurologists based on detailed examination results and were thoroughly documented in the medical record system.
Statistical analysis
The dataset was spatially divided into a training set and an independent validation set by applying EpiData 3.1 software to construct the database and performing a logistic test to confirm the correctness of the double-entry data. They were subsequently imported into SPSS 29.0. (SPSS Inc., Chicago, IL) and R (version 4.1.2; www.r-project.org) statistical software for data analysis.
Continuous variables were expressed as mean ± standard deviation; categorical variables were expressed as frequencies and percentages. When making comparisons between groups of measured data, independent samples t-tests or calibrated t-tests were used for one-way analyses to obtain more precise results, and χ2 tests were used to analyze count data. Include variables significant in univariate analysis in multivariate Logistic regression. Subsequently, variables with P < 0.05 in the multivariate analysis were retained to screen for the final independent predictors.
R statistical software (version 4.1.2; www.r-project.org) was used to construct the risk prediction model of nomogram based on the variables meaningful by multivariate analysis, and the degree of fit of the model to the data was evaluated by using the Hosmer-Lemeshow test and calibration curves. The caret package is used for data processing and model training operations, while the pROC package is employed to plot receiver operating characteristic curves and the area under the curve for evaluating model performance. The performance of the constructed risk prediction model can be assessed by evaluating the differentiation of the area under the curve (AUC) of the work characteristics of the subjects, sensitivity and specificity values, and the closer the value of these indexes is to 1, the better the performance of the model.
Meanwhile, the independent external validation cohort used the same sorting criteria to apply AUC, calibration curves, and other metrics for independent external validation of the model to verify the predictive ability of the prediction model so as to ensure its accuracy and reliability. In order to assess the value of the model in clinical practice, we used the method of decision curve analysis (DCA) to analyze the two models. When performing a two-sided test, if the p-value is less than 0.05, this indicates a statistically significant difference.
Ethics approval and consent to participate
The study followed the Declaration of Helsinki and was reviewed and approved by the Ethics Committee of the School of Nursing and Rehabilitation, Shandong University (Approval No. 2021-R-123). All participants were informed and gave informed consent to voluntarily participate in this study to ensure the fairness and reliability of the study.
Results
Basic characteristics of the training and validation sets
A total of 480 questionnaires were distributed between October 2022 and August 2023 for this study. Among the 480 older adult patients with heart failure, three patients in the end stage of heart failure disease were excluded; three patients with serious complications were not able to cooperate in completing the test; four patients were excluded due to refusal to participate or significant gaps in key information, a total of 470 valid questionnaires were recovered, of which the validity rate was as high as 98%. A total of 390 patients were assigned to the training set and 80 patients were assigned to the validation set based on patients in different hospital spaces. For detailed basic information, please refer to Table 1.
Table 1.
Baseline characteristics for training and validation sets
| Variables | Total (n = 470) |
Training set (n = 390) |
Validation set (n = 80) |
|---|---|---|---|
| Age, mean(SD) | 69.97(6.5) | 70.00(6.6) | 69.81(6.4) |
| Gender, n(%) | |||
| Male | 268(57.0) | 221(56.7) | 47(58.8) |
| Female | 202(43.0) | 169(43.3) | 33(41.2) |
| Education, n(%) | |||
| Illiterate | 95(20.2) | 77(19.7) | 18(22.5) |
| Primary school | 109(23.2) | 95(24.4) | 14(17.5) |
| Middle school | 133(28.3) | 108(27.7) | 25(31.3) |
| High school | 72(15.3) | 56(14.4) | 16(20.0) |
| College and above | 61(13.0) | 54(13.8) | 7(8.7) |
| Work status, n(%) | |||
| Incumbency | 78(16.6) | 57(14.6) | 21(26.3) |
| Retirement | 392(83.4) | 333(85.4) | 59(73.7) |
| Marital status, n(%) | |||
| Married | 398(84.7) | 327(83.8) | 71(88.8) |
| Unmarried | 72(15.3) | 63(16.2) | 9(11.2) |
| Smoking, n(%) | |||
| Never | 254(54.0) | 212(54.4) | 42(52.5) |
| Quit smoking | 160(34.1) | 129(33.1) | 31(38.8) |
| Smoking | 56(11.9) | 49(12.6) | 7(8.7) |
| Drinking, n(%) | |||
| Never | 252(53.6) | 209(53.6) | 42(53.8) |
| Quit drinking | 164(34.9) | 135(34.6) | 29(36.2) |
| Drinking | 54(11.5) | 46(11.8) | 8(10.0) |
| NYHA classification, n(%) | |||
| II | 208(44.3) | 184(47.2) | 24(30.0) |
| III | 180(38.3) | 135(34.6) | 45(56.3) |
| IV | 82(17.4) | 71(18.2) | 11(13.7) |
| NT-proBNP, mean(SD) | 3894.36(5928.6) | 3893.89(6068.2) | 3896.62(5229.6) |
| Chewing ability of teeth, n(%) | |||
| No problem | 285(60.6) | 227(58.2) | 58(72.5) |
| Yes, but be careful | 116(24.7) | 101(25.9) | 15(18.8) |
| Not at all | 69(14.7) | 62(15.9) | 7(8.7) |
| LVEF, mean(SD) | 0.47(0.2) | 0.47(0.2) | 0.47(0.2) |
| Nutritional status, mean(SD) | 10.88(2.7) | 10.90(2.7) | 10.78(2.6) |
| Loneliness, mean(SD) | 14.84(5.2) | 15.02(5.2) | 13.95(4.6) |
| Sleep status, mean(SD) | 6.20(5.7) | 6.36(5.9) | 5.40(4.5) |
| Depression, mean(SD) | 1.40(1.5) | 1.43(1.6) | 1.21(1.2) |
| Social support, mean(SD) | 35.32(6.9) | 35.01(6.6) | 36.85(8.0) |
| Activities of daily living, mean(SD) | 90.56(15.9) | 90.45(16.2) | 91.09(14.1) |
Abbreviations: NYHA classification New York Heart association classification, LVEF, left ventricular ejection fraction
Univariate and multivariate analysis of the training set
Univariate and multivariate logistic regression analysis of all variables in the development cohort showed that the differences between the two groups were statistically significant (P < 0.05) in terms of Age, NYHA classification, chewing ability of teeth, nutritional status, loneliness, and sleep status. These six factors are considered to be independent predictors of the development of cognitive frailty in older adult patients with heart failure (Tables 2 and 3).
Table 2.
Univariate analysis of the training set (n = 390)
| Variables | Cognitive frailty(n = 183) | Non-cognitive frailty(n = 207) | t/χ2 | P Value |
|---|---|---|---|---|
| Age (years) | 71.0 ± 7.1 | 69.1 ± 6.2 | -2.807 | 0.005 |
| Gender | 0.306 | 0.609 | ||
| Male | 101(55.2) | 120(58.0) | ||
| Female | 82(44.8) | 87(42.0) | ||
| Education | 8.144 | 0.087 | ||
| Illiterate | 31(16.9) | 46(22.2) | ||
| Primary school | 50(27.3) | 45(21.7) | ||
| Middle school | 57(31.1) | 51(24.6) | ||
| High school | 27(14.8) | 29(14.0) | ||
| College and above | 18(9.8) | 36(17.4) | ||
| Work status | 4.950 | 0.031 | ||
| Incumbency | 19(10.4) | 38(18.4) | ||
| Retirement | 164(89.6) | 169(81.6) | ||
| Marital status | 4.206 | 0.053 | ||
| Married | 146(79.8) | 181(87.4) | ||
| Unmarried | 37(20.2) | 26(12.6) | ||
| Smoking | 6.619 | 0.036 | ||
| Never | 108(59.0) | 104(50.2) | ||
| Quit smoking | 60(32.8) | 69(33.3) | ||
| Smoking | 15(8.2) | 34(16.4) | ||
| Drinking | 7.654 | 0.020 | ||
| Never | 106(57.9) | 103(49.8) | ||
| Quit drinking | 64(35.0) | 71(34.3) | ||
| Drinking | 13(7.1) | 33(15.9) | ||
| NT-proBNP | 5202.8 ± 7714.2 | 2736.8 ± 3753.9 | -3.932 | < 0.001 |
| LVEF | 0.46 ± 0.16 | 0.49 ± 0.15 | 1.978 | 0.049 |
| NYHA classification | 35.428 | < 0.001 | ||
| II | 59(32.2) | 125(60.4) | ||
| III | 74(40.4) | 61(29.5) | ||
| IV | 50(27.3) | 21(10.1) | ||
| Chewing ability of teeth | 32.974 | < 0.001 | ||
| No problem | 81(44.3) | 146(70.5) | ||
| Yes, but be careful | 56(30.6) | 45(21.7) | ||
| Not at all | 46(25.1) | 16(7.70) | ||
| Nutritional status | 10.0 ± 2.85 | 11.7 ± 2.2 | 6.404 | < 0.001 |
| Loneliness | 17.0 ± 5.3 | 13.3 ± 4.5 | -7.227 | < 0.001 |
| Sleep status | 8.2 ± 6.5 | 4.7 ± 4.8 | -5.901 | < 0.001 |
| Depression | 2.0 ± 1.7 | 0.97 ± 1.21 | -6.420 | < 0.001 |
| Social support | 33.2 ± 6.4 | 36.6 ± 6.4 | 5.195 | < 0.001 |
| Activities of daily living | 84.5 ± 20.0 | 95.7 ± 9.2 | 7.009 | < 0.001 |
Abbreviations: NYHA classification, New York Heart association classification, LVEF, left ventricular ejection fraction
Table 3.
Multivariate analysis of the training set (n = 390)
| Variables | P Value | OR | 95%CI | |
|---|---|---|---|---|
| Age (years) | - | 0.042 | 1.039 | 1.001-1.078 |
| NYHA classification | Ⅱ | |||
| Ⅲ | 0.007 | 2.038 | 1.219-3.410 | |
| Ⅳ | 0.001 | 2.971 | 1.518-5.815 | |
| Chewing ability of teeth | No problem | |||
| Yes, but be careful | 0.613 | 1.154 | 0.661-2.015 | |
| Not at all | 0.017 | 2.424 | 1.175-5.003 | |
| Nutritional status | - | 0.001 | 0.838 | 0.758-0.926 |
| Loneliness | - | < 0.001 | 1.109 | 1.053-1.167 |
| Sleep status | - | 0.008 | 1.062 | 1.016-1.110 |
Abbreviations: NYHA classification, New York Heart association classification
Development of the nomogram risk prediction model
Based on the results of multivariate logistic regression, the six identified variables were plotted in a nomogram risk prediction model (Fig. 1). We conducted goodness-of-fit tests on the established model and found that the Hosmer-Lemeshow(HL) test yielded a P value of 0.585. Generally, a P value greater than or equal to 0.05 indicates that the differences between predicted and actual values are not significant following the HL test. Additionally, a calibration curve was plotted for this study, as shown in Fig. 2. The results collectively indicate that the model exhibits good fit, confirming the reliability and validity of the established predictive model.
Fig. 1.
Nomogram model for predicting the development of cognitive frailty in older adult patients with heart failure
Fig. 2.
Performance evaluation for predicting the risk of developing cognitive frailty in older adult patients with heart failure; (A) Receiver operating characteristic in the training set; (B) Receiver operating characteristic in the validation set; (C) Calibration curves in the training set; (D) Calibration curves in the validation set; (E) Decision curve analysis in the training set; (F) Decision curve analysis in the validation set
The nomogram risk prediction model consists of six independent predictors on variable axes, with a score scale at the top, and a total score scale and risk probability scale at the bottom. The corresponding points of each variable corresponded to different scores, and the scores were summed to obtain a total score, which corresponded to the risk probability of developing cognitive frailty.
Independent external validation of the nomogram risk prediction model for cognitive frailty in older adult patients with heart failure
The predictive effectiveness of the nomogram prediction model is shown in Fig. 2. The predictive ability AUC of the model development cohort was 0.807 (95%CI = 0.764-0.851,P<0.001) (Fig. 2A). In contrast, the validation cohort had an AUC of 0.770 (95%CI = 0.667-0.874,P<0.001) for predictive power (Fig. 2B). The calibration curves of the model showed that the predicted values were generally in good agreement with the measured values (Fig. 2C, D). To determine the value of the clinical application of the nomogram model, decision curve plots were drawn to assess the net clinical benefit of using the nomogram model to predict the risk of developing cognitive frailty in older adult patients with chronic heart failure, and most of the prediction threshold probability curves were above the two extreme lines in the range of 0 ~ 1 threshold probabilities, indicating that the use of the model could result in a net benefit to the clinical workup (Fig. 2E, F).
Discussion
In this study, we comprehensively analyzed for the first time the current status and factors influencing cognitive frailty in older adult patients with heart failure and developed the first screening tool for cognitive frailty. The results of the study showed that the overall prevalence of cognitive frailty in older adult patients with heart failure was as high as 44.5%. This suggests that there is an urgent need for clinical caregivers to focus on this population of older adult patients with heart failure and to develop timely intervention strategies to reduce the incidence of cognitive frailty.
In addition, there are more factors affecting the occurrence of cognitive frailty in older adult patients with heart failure: the various physiological functions of heart failure patients will gradually weaken with age, while the hippocampus and the cerebral cortex accelerate atrophy, leading to a continuous decline in the cognitive function of the body, thus increasing the risk of cognitive frailty [27]. The higher the NYHA classification, the lower the cardiac output, and decreased cardiac output has been shown to not only be a potential cause of sarcopenia, but also to contribute to cognitive impairment by directly reducing cerebral blood flow [28, 29]. Patients with malnutrition have a high prevalence of cognitive frailty, and some findings have shown a negative correlation between nutritional status and cognitive frailty [30]. In a study that included 4093 community-dwelling older adults, it was supported that interventions on sleep can help prevent the onset of cognitive frailty in older adult [31]. Some scholars and academics explored the relationship between dental status and cognitive frailty using the NHANES database and showed that older adult with more teeth had a lower risk of cognitive frailty [32].
We screened six independent predictor variables for cognitive frailty in older adult patients with heart failure and constructed a predictive model based on these variables for assessing the risk of developing cognitive frailty. The nomogram is essentially a visual representation of a complex mathematical formula. One of its significant strengths is its ability to incorporate patient- and disease-specific characteristics to provide accurate risk estimates for individuals. By incorporating relevant continuous and categorical variables into the model and clearly demonstrating on a scale the extent to which each variable contributes to the predicted outcome, the nomogram model achieves a high degree of visualization of the prediction process.
In addition, the six clinical predictors on which the model is based are commonly used in the daily clinical assessment of older adult patients with heart failure, which are not only easy to access, but also simple and clear in the assessment process. This ensures the clinical utility and feasibility of the model and provides a strong guarantee for the early prediction of cognitive frailty risk. In addition, in the Nomogram-based risk factor prediction model, targeted interventions are needed for the above factors by clinical healthcare professionals and social workers. NYHA classification is often assessed at the time of admission to the hospital, and this classification may improve during hospitalization, requiring doctors and nurses to pay attention to changes in cardiac function in older adult patients with heart failure. Nutritional status, loneliness, and sleep are often prone to change compared to age, dental chewing ability, and NYHA classification, so it is important to individualize medical and nursing interventions for each patient’s condition to ensure that they receive the best possible outcome, giving older adult patients with heart failure the best possible nutritional eating advice, encouraging patients to engage in social interaction, and adopting some sleep aids are some of the precise interventions to provide more scientific decision-making basis for the comprehensive health management of patients.
In order to validate the extrapolation and predictive stability of the constructed nomogram prediction model for the risk of cognitive frailty in older adult patients with heart failure, we adopted a spatially independent external validation strategy to collect and validate the data from two hospitals in particular. In this process, AUC serves as an important yardstick to measure the overall discriminative ability of the model, and the magnitude of its value directly reflects the predictive performance of the model. In general, the higher the AUC value, the stronger the prediction accuracy of the model. The results of this study showed that the AUC value of the model exceeded the threshold of 0.75 in both the training and validation groups, which strongly demonstrated the model’s excellent discriminative ability and predictive effectiveness in distinguishing between cognitive frailty and non-cognitive frailty. Further, we also utilized the calibration curves to assess the agreement between the model predictions and the actual clinical observations. The results showed that the model predictions were highly consistent with the actual situation, which not only improved the accuracy of the predictions but also enhanced their rigor. In conclusion, the nomogram prediction model not only has excellent predictive performance, but also has high clinical utility value, which makes it an important tool for assessing the risk of cognitive frailty in older adult patients with heart failure.
This study has several limitations. Considering the cross-sectional study design used in this study, the causal relationship between variables could not be inferred, therefore, more in-depth longitudinal studies are needed to explore the factors influencing the cognitive frailty of hospitalized older adult patients with chronic heart failure; in addition, although independent external validation was conducted in this study, future expansion of the sample size and conduct multicenter external validation to ensure the reliability of the model with the aim of providing a more reliable and stable risk screening tool for the clinic and society.
Conclusion
The high prevalence of cognitive frailty in older adult patients with heart failure is not an optimistic situation. This suggests that there is an urgent need for clinical medical and nursing staff to pay attention to the prevalence of cognitive frailty in this population and to develop timely intervention strategies to reduce the incidence of cognitive frailty. Through univariate and multivariate regression analyses, it was concluded that the occurrence of cognitive frailty in older adult patients with heart failure is closely related to age, NYHA classification, chewing ability of teeth, nutritional status, loneliness, and sleep status. Based on the independent variables, we developed and independently validated a new nomogram risk prediction model, which can make the prediction more graphical and visual, conveniently, quickly and effectively assess the probability of the occurrence of cognitive frailty risk in older adult patients with heart failure, and provide a new tool for clinical caregivers to improve the quality of care, and provide a powerful decision-making support for the development and implementation of early interventions.
Authors' contributions
Jian Liu, Shengjia Xu and Xiaorong Luan designed the research; Zhiwei Wang and Xueqing Song collected the data; Zongke Long, Shengjia Xu, Jian Liu and Zhiwei Wang analyzed the data; Jian Liu wrote the manuscript; Shengjia Xu, Zhiwei Wang reviewed and edited the manuscript; Xiaorong Luan had primary responsibility for final content. All authors have read and agreed to the published version of the manuscript.
Funding
This study was sponsored by the National Natural Science Foundation of China (NO. 72474121).
Data availability
The data that support the findings of this study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
The study followed the Declaration of Helsinki and was reviewed and approved by the Ethics Committee of the School of Nursing and Rehabilitation, Shandong University (Approval No. 2021-R-123). All participants were informed and gave informed consent to voluntarily participate in this study to ensure the fairness and reliability of the study.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Jian Liu and Shengjia Xu contributed equally to this work.
References
- 1.Kittipibul V, Lam CSP. Heart failure with preserved ejection fraction and atrial fibrillation: epidemiology, pathophysiology, and diagnosis interplay. Heart Fail Rev. 2025;30(4):697–705. [DOI] [PubMed] [Google Scholar]
- 2.Bozkurt B, Ahmad T, Alexander K, Baker WL, Bosak K, Breathett K, Carter S, Drazner MH, Dunlay SM, Fonarow GC, et al. HF STATS 2024: Heart failure epidemiology and outcomes statistics an updated 2024 Report from the heart failure society of america. J Card Fail. 2025;31(1):66–116. [DOI] [PubMed] [Google Scholar]
- 3.Becher PM, Lund LH, Coats AJS, Savarese G. An update on global epidemiology in heart failure. Eur Heart J. 2022;43(32):3005–7. [DOI] [PubMed] [Google Scholar]
- 4.Sunayama T, Matsue Y, Dotare T, Maeda D, Iso T, Morisawa T, Saitoh M, Yokoyama M, Jujo K, Takahashi T, et al. Multidomain Frailty as a therapeutic target in elderly patients with heart failure. Int Heart J. 2022;63(1):21–839. [DOI] [PubMed] [Google Scholar]
- 5.Denfeld QE, Winters-Stone K, Mudd JO, Gelow JM, Kurdi S, Lee CS. The prevalence of frailty in heart failure: A systematic review and meta-analysis. Int J Cardiol. 2017;236:283–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kwok CS, Zieroth S, Van Spall HGC, Helliwell T, Clarson L, Mohamed M, Mallen C, Duckett S, Mamas MA. The hospital frailty risk score and its association with in-hospital mortality, cost, length of stay and discharge location in patients with heart failure short running title: Frailty and outcomes in heart failure. Int J Cardiol. 2020;300:184–90. [DOI] [PubMed] [Google Scholar]
- 7.Zhang H, Jie Y, Sun Y, Wang X, Gong D, Fan Y. Association of cognitive impairment with mortality and readmission in patients with heart failure: a meta-analysis. Curr Prob Cardiol. 2022;47(12):101354. [DOI] [PubMed] [Google Scholar]
- 8.Wu Y, Chen L, Zhong F, Zhou K, Lu C, Cheng X, Wang S. Cognitive impairment in patients with heart failure: molecular mechanism and therapy. Heart Fail Rev. 2023;28(4):807–20. [DOI] [PubMed] [Google Scholar]
- 9.Kelaiditi E, Cesari M, Canevelli M, Van Abellan G, Ousset PJ, Gillette-Guyonnet S, Ritz P, Duveau F, Soto ME, Provencher V, et al. Cognitive frailty: rational and definition from an (I.A.N.A./I.A.G.G.) International consensus group. J Nutr health aging. 2013;17(9):726–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Dent E, Morley JE, Cruz-Jentoft AJ, Woodhouse L, Rodriguez-Manas L, Fried LP, Woo J, Aprahamian I, Sanford A, Lundy J, et al. Physical frailty: ICFSR international clinical practice guidelines for identification and management. J Nutr Health Aging. 2019;23(9):771–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Gale C, Ritchie SJ, Starr JM, Deary IJ. Physical frailty and decline in general and specific cognitive abilities: the Lothian Birth Cohort 1936. J Epidemiol Commun H. 2020;74(2):108–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Malmstrom TK, Morley JE. Frailty and cognition: linking two common syndromes in older persons. J Nutr Health Aging. 2013;17(9):723–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Fougère B, Delrieu J, Del Campo N, Soriano G, Sourdet S, Vellas B. RETRACTED: cognitive frailty: mechanisms, tools to measure, prevention and controversy. Clin Geriatr Med. 2017;33(3):339–55. [DOI] [PubMed] [Google Scholar]
- 14.Seo EJ, Son Y. The prevalence of cognitive frailty and its association with sleep duration and depression among older adults with heart failure. Clin Gerontologist. 2024;47(3):416–25. [DOI] [PubMed] [Google Scholar]
- 15.Xu J, Xiang L, Zhang H, Sun X, Xu D, Wu D, Chen C, Zhang Y, Gu Z. Prevalence and modifiable risk factors of cognitive frailty in patients with chronic heart failure in China: a cross-sectional study. Bmc Cardiovasc Disor 2024;24(1). [DOI] [PMC free article] [PubMed]
- 16.Miao X, Guo Y, Ding L, Xu X, Zhao K, Zhu H, Chen L, Chen Y, Zhu S, Xu Q. A dynamic online nomogram for predicting the heterogeneity trajectories of frailty among elderly gastric cancer survivors. Int J Nurs Stud. 2024;153:104716. [DOI] [PubMed] [Google Scholar]
- 17.Gou D, Min C, Peng X, Wu H, Zhang L, Chen Y, Tao M. Associating factors of cognitive frailty among older people with chronic heart failure: Based on LASSO-logistic regression. J Adv Nurs. 2025;81(3):1399–411. [DOI] [PubMed] [Google Scholar]
- 18.Riley RD, Ensor J, Snell KIE, Harrell FE, Martin GP, Reitsma JB, Moons KGM, Collins G, van Smeden M. Calculating the sample size required for developing a clinical prediction model. Bmj-Brit Med J. 2020;368:m441. [DOI] [PubMed] [Google Scholar]
- 19.Rubenstein LZ, Harker JO, Salva A, Guigoz Y, Vellas B. Screening for undernutrition in geriatric practice: developing the short-form mini-nutritional assessment (MNA-SF). J Gerontol A-Biol. 2001;56(6):M366–72. [DOI] [PubMed] [Google Scholar]
- 20.Soldatos CR, Dikeos DG, Paparrigopoulos TJ. Athens insomnia scale: validation of an instrument based on ICD-10 criteria. J Psychosom Res. 2000;48(6):555–60. [DOI] [PubMed] [Google Scholar]
- 21.Russell D, Peplau LA, Ferguson ML. Developing a measure of loneliness. J Pers Assess. 1978;42(3):290–4. [DOI] [PubMed] [Google Scholar]
- 22.Hoyl MT, Alessi CA, Harker JO, Josephson KR, Pietruszka FM, Koelfgen M, Mervis JR, Fitten LJ, Rubenstein LZ. Development and testing of a five-item version of the geriatric depression scale. J Am Geriatr Soc. 1999;47(7):873–8. [DOI] [PubMed] [Google Scholar]
- 23.Cheng Y, Liu C, Mao C, Qian J, Liu K, Ke G. Social support plays a role in depression in Parkinson’s disease: a cross-section study in a Chinese cohort. Parkinsonism Relat D. 2008;14(1):43–5. [DOI] [PubMed] [Google Scholar]
- 24.Malmstrom TK, Miller DK, Morley JE. A comparison of four frailty models. J Am Geriatr Soc. 2014;62(4):721–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.van Kan GA, Rolland YM, Morley JE, Vellas B. Frailty: Toward a clinical definition. J Am Med Dir Assoc. 2008;9(2):71–2. [DOI] [PubMed] [Google Scholar]
- 26.Nasreddine ZS, Phillips NA, Bédirian V, Charbonneau S, Whitehead V, Collin I, Cummings JL, Chertkow H. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005;53(4):695–9. [DOI] [PubMed] [Google Scholar]
- 27.Qiu Y, Li G, Wang X, Zheng L, Wang C, Wang C, Chen L. Prevalence of cognitive frailty among community-dwelling older adults: a systematic review and meta-analysis. Int J Nurs Stud. 2022;125:104112. [DOI] [PubMed] [Google Scholar]
- 28.Diniz BS, Lima-Costa MF, Peixoto SV, Firmo JOA, Torres KCL, Martins-Filho OA, Teixeira-Carvalho A, Grady J, Kuchel GA, Castro-Costa E. Cognitive Frailty is associated with elevated proinflammatory markers and a higher risk of mortality. Am J Geriat Psychiat. 2022;30(7):825–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kocagoncu E, Nesbitt D, Emery T, Hughes LE, Henson RN, Rowe JB. Neurophysiological and Brain structural markers of cognitive frailty differ from Alzheimer’s disease. J Neurosci. 2022;42(7):1362–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kwan RYC, Leung AYM, Yee A, Lau LT, Xu XY, Dai DLK. Cognitive frailty and its association with nutrition and depression in community-dwelling older people. J Nutr Health Aging. 2019;23(10):943–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Zhao Y, Lu Y, Zhao W, Wang Y, Ge M, Zhou L, Yue J, Dong B, Hao Q. Long sleep duration is associated with cognitive frailty among older community-dwelling adults: results from West China Health and Aging Trend study. Bmc Geriatr 2021;21(1). [DOI] [PMC free article] [PubMed]
- 32.Zhang XM, Wu X, Chen W. the association between number of teeth and cognitive frailty in older adults: a cross-sectional study. J Nutr health aging. 2022;26(5):430–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available from the corresponding author on reasonable request.