System and method for analyzing and predicting malnutrition of elderly patients with mild cognitive impairment

Abstract

The invention discloses a system and a method for analyzing and predicting malnutrition of patients with elderly mild cognitive impairment, belongs to the technical field of malnutrition analysis and prediction, incorporates a more comprehensive and systematic index system, and covers potential risk factors, traditional risk factors and laboratory data of the patients with elderly mild cognitive impairment. The comprehensive breakthrough of the research perspective is realized, and the limitation defect of the prior art is broken through. The decision tree model and the logistic regression model are comprehensively compared, a precise malnutrition risk prediction model for the elderly MCI patient is constructed, a reference with practical value is provided for clinical prevention and treatment work, and optimization and perfection of a malnutrition prevention and treatment strategy of the elderly mild cognitive impairment patient are expected to be promoted. According to the method, the defects that in the prior art, analysis of the malnutrition of the elderly mild cognitive impairment patient has limitation, and a traditional prediction model is insufficient in decision making and pertinence aspects of the malnutrition of the elderly mild cognitive impairment patient are effectively overcome.

Description

Technical Field

[0001] The present invention belongs to the technical field of analyzing and predicting malnutrition, and specifically relates to a system and method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment. Background Art

[0002] In the field of prevention and treatment of mild cognitive impairment (MCI), diet-based non-drug interventions have shown application potential in the prevention and management of MCI. Malnutrition has a profound impact on elderly patients with mild cognitive impairment throughout all stages of the disease development.

[0003] Although previous studies under the existing technical solution with patent publication number "CN113628750B" have identified malnutrition risk factors for some elderly patients with mild cognitive impairment, there are limitations, and the traditional prediction models used are insufficient in decision-making and targeting for malnutrition in elderly patients with mild cognitive impairment. Summary of the Invention

[0004] In order to address the defects in the existing technology, the present invention proposes a system and method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment. The present invention effectively avoids the limitations of the existing technology in analyzing malnutrition in elderly patients with mild cognitive impairment and the deficiencies of traditional prediction models in decision-making and targeting in the face of malnutrition in elderly patients with mild cognitive impairment.

[0005] The present invention utilizes the following technical solutions.

[0006] A method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment, comprising: Convenience sampling was used to select data from corresponding hospitals for cross-sectional analysis, which specifically included: Step 1: Determine the analysis object for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment; Step 2: Perform group nutritional screening on the subjects to be analyzed to predict malnutrition in elderly patients with mild cognitive impairment; Step 3: Perform malnutrition risk prediction for the analysis object of analyzing and predicting malnutrition in elderly patients with mild cognitive impairment.

[0007] Furthermore, in step 1, the inclusion criteria for the analysis subjects for predicting malnutrition in elderly patients with mild cognitive impairment are determined as follows: (1) Age ≥ 60 years; (2) diagnosed with MCI; (3) self-reported no visual or hearing impairment; (4) Able to sign informed consent.

[0008] Furthermore, in step 1, the operational criteria used for screening to confirm MCI include: i. The subject or informant reported a relative decline in cognitive function in the past year; ii. The subject has objective cognitive impairment that is inconsistent with age and education level; iii. The ability to carry out activities of daily living is intact; iv. No dementia; Subjects were excluded from the diagnosis of MCI if they had the following conditions: (1) A history of serious neurological, mental or other diseases that may affect brain function; (2) Taking any medication that may impair or improve cognitive function in the past 6 months; (3) Participated in another training related to cognitive function within 1 year.

[0009] Furthermore, in step 1, the subject may have objective cognitive impairment that is inconsistent with age and education level, including: Those with primary school education had a score of less than 18 points on the Montreal Cognitive Assessment Basic Version, those with secondary school education had a score of less than 21 points, and those with higher education had a score of less than 23 points.

[0010] Furthermore, in step 1, the ability to perform daily living activities is intact when the subject's Lawton-Brody score for daily living activities is ≤16 points.

[0011] Furthermore, in step 2, the 2002 version of the nutritional risk screening tool is used to perform group nutritional screening on the analysis subjects for predicting malnutrition in elderly patients with mild cognitive impairment.

[0012] Furthermore, in step 2, the NRS-2002 scale of the 2002 version of the nutritional risk screening tool is used to assess three dimensions: impaired nutritional status, disease severity, and age. The sum of the scores of each dimension constitutes the total screening score, that is, an NRS-2002 total score below 3 points indicates good nutritional status, while 3 points or above indicates malnutrition.

[0013] Furthermore, step 3 specifically includes: Step 3-1: Identify demographic characteristics and lifestyle factors for malnutrition risk prediction; Step 3-2: Identify physiological indicators for malnutrition risk prediction; Step 3-3: Perform statistical analysis and prediction on physiological indicators, demographic characteristics and lifestyle factors.

[0014] Furthermore, in step 3-1, the demographic characteristics include age, gender, education level, marital status, family income, residential address, medical insurance, polypharmacy, and number of chronic comorbidities of the analysis subjects for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment.

[0015] Furthermore, in step 3-1, the lifestyle factors include analyzing and predicting the malnutrition of the elderly patients with mild cognitive impairment by analyzing the appetite, daily eating habits, regular physical activities, sleep quality, self-care ability, caregiver efficacy, and dietary diversity information of the analysis subject; Appetite was assessed using the simplified nutritional appetite questionnaire; Daily eating habits and regular physical activity were dichotomous variables, defined as follows: daily eating habits refer to whether or not eating regularly; regular eating refers to regular meal times and frequencies, and eating at least three meals a day on time is considered regular eating; regular physical activity refers to walking, square dancing, or cycling for 150 minutes or more per week in the past month; Sleep quality was assessed using the Pittsburgh Sleep Quality Index; Self-care ability was measured using the Activities of Daily Living Scale, which assesses basic activities of daily living and instrumental activities of daily living; Caregiver efficacy was measured using the General Self-Efficacy Scale for Primary Caregivers; Dietary diversity was calculated based on the dietary diversity score.

[0016] Furthermore, in step 3-2, the physiological indicators include analyzing and predicting malnutrition in the elderly patients with mild cognitive impairment by analyzing the subject's body mass index, oral health status, overall cognitive function, disorientation, immediate memory impairment, attention deficit, retrograde amnesia, language disorder, neuropsychiatric symptoms, dysphagia, hemoglobin, blood potassium, blood sodium, cholesterol, and albumin; BMI is an internationally recognized indicator for measuring body fat content and health status. The calculation formula is: BMI = weight (kg) / height² (m²); Oral health status was assessed using an oral health assessment tool; Global cognitive function was assessed using the Mini-Mental State Examination; Neuropsychiatric symptoms were measured using the Neuropsychiatric Rating Scale; Dysphagia was assessed using the Kubota water drinking test to assess the patient's swallowing function; Biochemical indices such as serum potassium, serum sodium, cholesterol, albumin, and hemoglobin were collected through the hospital electronic medical record system.

[0017] Furthermore, in step 3-3, the following test data were collected for all subjects: nutritional status, age, gender, education level, marital status, family income, residential address, medical insurance, polypharmacy, number of chronic diseases, appetite, daily eating habits, regular physical activity, sleep status, self-care ability, caregiver efficacy, dietary diversity, BMI, oral health status, overall cognitive function, disorientation, immediate memory impairment, attention deficit, retrograde amnesia, language disorder, neuropsychiatric symptoms, dysphagia, hemoglobin, blood potassium, blood sodium, cholesterol, and albumin; The test data were statistically analyzed using R-4.4.2 and SPSS27.0 software. If the test data followed a normal distribution, they were expressed as mean ± standard deviation (x ± s), and an independent sample t-test was used. If the test data followed a non-normal distribution, they were expressed as the median of the test data, and the Mann-Whitney U test was used for the median of the test data. If the test data were expressed as the number of cases, a chi-squared test was used. Lasso regression was used to reduce the dimensionality of the monitoring data that passed the test and identify the influencing factors of malnutrition in elderly patients with mild cognitive impairment. In addition, binary logistic regression and decision tree were used to establish a clinical prediction model for the monitoring data as an evaluation model. The Hosmer-Lemeshow test was used to evaluate the goodness of fit of the model and to draw a calibration curve. The area under the receiver operating characteristic curve was used to determine the discrimination of the evaluation model.

[0018] A system for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment, comprising: Determine module 1, which is used to determine the analysis object for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment; Determine a second module for performing group nutrition screening on analysis subjects for predicting malnutrition in elderly patients with mild cognitive impairment; The prediction module is used to perform malnutrition risk prediction for an analysis object that analyzes and predicts malnutrition in elderly patients with mild cognitive impairment.

[0019] The beneficial effects of the present invention are that, compared with the prior art, the technical effects of the present invention include: This study incorporates a more comprehensive and systematic indicator system, encompassing potential risk factors, traditional risk factors, and laboratory data for elderly patients with mild cognitive impairment. This represents a comprehensive breakthrough in research perspectives and overcomes the limitations of existing technologies. By comprehensively comparing decision tree models with logistic regression models, a precise malnutrition risk prediction model for elderly patients with MCI has been constructed. This provides a valuable reference for clinical prevention and treatment, and is expected to promote the optimization and improvement of malnutrition prevention and treatment strategies for elderly patients with mild cognitive impairment. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a flow chart of the method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to the present invention; Figure 2 This is a partial structural diagram of the system for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to the present invention; Figure 3 Schematic diagram of the decision tree model analysis of malnutrition risk factors in elderly patients with mild cognitive impairment according to the present invention; Figure 4 Schematic diagram of the receiver operating characteristic curves of the logistic regression and decision tree models described in the present invention. DETAILED DESCRIPTION

[0021] Although previous studies using existing technical solutions have identified malnutrition risk factors for some elderly patients with mild cognitive impairment, they have limitations, and the traditional prediction models used are insufficient in decision-making and targeting.

[0022] The decision tree model, with its ability to automatically capture complex data patterns and intuitively present data-variable relationships, can provide healthcare professionals with more efficient and accurate decision support.

[0023] This study aims to investigate the risk factors and predictive value of malnutrition in elderly individuals with mild cognitive impairment (MCI). Given the significant impact of malnutrition on health status, quality of life, and medical burden for this population, the results of this study aim to provide a basis for targeted interventions for this vulnerable population.

[0024] The method presented in this paper primarily utilizes a convenience sampling, cross-sectional study. First, data were collected through a literature review and consultation with in-hospital experts. The collected data covered 31 influencing factors, including age, gender, education level, marital status, household income, residential address, medical insurance, polypharmacy, number of comorbid chronic diseases, appetite, daily dietary habits, regular physical exercise, sleep patterns, self-care ability, caregiver effectiveness, dietary diversity, body mass index (BMI), oral health, global cognitive function, disorientation, immediate memory impairment, attention deficit, retrograde amnesia, language impairment, neuropsychiatric symptoms, dysphagia, hemoglobin, serum potassium, serum sodium, cholesterol, and albumin. A questionnaire was administered to elderly patients with MCI aged 60 years and older at relevant hospitals (e.g., two tertiary hospitals), and blood samples were collected and analyzed. Patients were divided into a malnutrition risk group and a normal group. Univariate analysis, lasso regression, decision tree analysis, and logistic regression analysis were used to screen and analyze these factors to construct a malnutrition prediction model for this patient population.

[0025] Validation of the method presented in this study revealed that a total of 170 valid questionnaires were collected from elderly patients with MCI, including 66 in the malnutrition risk group and 104 in the normal group. Univariate analysis revealed significant differences in 11 variables (P < 0.05). Lasso regression identified seven factors significantly influencing malnutrition (P < 0.05). The areas under the receiver operating characteristic curves (ROC-AUCs) for the logistic regression model and the decision tree model were 0.819 (95% confidence interval: 0.752-0.887) and 0.824 (95% confidence interval: 0.753-0.895), respectively, indicating that the decision tree model had slightly superior predictive performance.

[0026] The method presented in this paper concluded that BMI, oral health status, number of comorbid chronic diseases, hemoglobin levels, and appetite significantly increase the risk of malnutrition in elderly patients with MCI. Understanding these factors can help healthcare providers more accurately assess the risk of malnutrition in elderly patients with MCI, thereby assisting in developing personalized prevention and intervention strategies.

[0027] To make the objectives, technical solutions, and advantages of the present invention more clear, the following will be combined with the accompanying drawings in the embodiments of the present invention to clearly and completely express the technical solutions of the present invention. The embodiments expressed in this application are only some embodiments of the present invention, not all embodiments. Based on the spirit of the present invention, other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0028] like Figure 1 As shown, the method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to the present invention comprises: Mild cognitive impairment (MCI) is an intermediate condition between normal aging and dementia. Globally, it affects 10%-20% of adults aged 65 and over. In China, the prevalence reaches 15.54% among adults aged 60 and over, and continues to rise. While only 1%-2% of cognitively healthy older adults will develop dementia each year, over 50% of patients with MCI will progress to dementia within 5 years; however, 14%-55% of patients can be reversible with effective management. Growing evidence emphasizes the important role of non-pharmacological interventions, particularly dietary strategies, in the prevention and care of MCI. A meta-analysis published in the medical journal The Lancet suggests that lifestyle changes can prevent or delay 35%-40% of Alzheimer's disease-related dementia. The 2022 International Task Force on Nutrition and Dementia Prevention noted that dietary nutrition, as a key environmental factor, influences cognitive trajectory through the gut microbiome-gut-brain axis and multiple biological pathways (behavioral, genetic, immune, metabolic, and vascular).

[0029] Elderly patients with mild cognitive impairment face both cognitive decline and widespread malnutrition, a core factor affecting their quality of life, disease progression, and prognosis. A longitudinal cohort study in Colombia showed that the overall prevalence of malnutrition among the Alzheimer's disease (AD) population was 28.7%, while this rate was 54.1% among female patients with MCI and early AD attending the National Geriatrics Center's memory clinic. An observational study in China showed that the prevalence of malnutrition among home-based AD patients was 51.67%, while a cross-sectional study showed that the rate was 42.50% among elderly hospitalized AD patients. Although data vary across regions, the problem cannot be ignored. An observational study found that malnutrition was most common among patients with MCI, and cohort studies have shown that patients with MCI and AD have higher resting energy expenditure than controls. Therefore, early and accurate identification of modifiable factors of malnutrition and proactive intervention are crucial to improving patients' nutritional status and prognosis.

[0030] From a practical perspective, malnutrition has a multifaceted negative impact on elderly patients with mild cognitive impairment. On the one hand, it leads to a further decline in physical function and further limitations in daily activities. On the other hand, malnutrition weakens the immune system, significantly increasing the risk of infection. Most importantly, it accelerates cognitive deterioration and accelerates progression to dementia. Malnutrition impacts Alzheimer's disease throughout the disease course. In 2015, the European Society for Clinical Nutrition and Metabolism (ESPEN) guidelines stated that the most prominent nutrition-related issues in patients with dementia are weight loss and malnutrition. These two changes can occur at all stages of the disease and become increasingly pronounced with progression. Studies have shown that unexplained weight loss in the elderly over a short period of time is considered a preclinical manifestation of Alzheimer's disease. Patients with mild to moderate Alzheimer's disease may experience a 30%-40% weight loss. A prospective multicenter study showed that malnourished patients had a higher rate of disease progression compared with preclinical and clinical Alzheimer's patients with normal nutritional status. Malnutrition was an independent risk factor for disease progression (odds ratio: 2.4, 95% confidence interval: 1.1-5.1). A 6.5-year prospective cohort study demonstrated that rapid weight loss (≥5 kg within 6 months) in patients with AD can be a predictor of mortality. Therefore, the impact of malnutrition on AD spans the entire disease process, from onset to progression and prognosis. Early nutritional intervention is crucial for preventing or delaying the onset of malnutrition and even AD-related cognitive impairment.

[0031] It is necessary to clarify the causes of malnutrition at different stages of the disease and implement corresponding nutritional intervention measures. On the one hand, these intervention measures should solve the existing malnutrition problem, while promoting the independence of patients in daily life as much as possible. Nutritional intervention should follow the principles of "early, coordinated, comprehensive and long-term". It emphasizes the importance of lifestyle adjustments such as diet and nutrition as a "zero-level prevention" strategy for AD-related cognitive impairment, as well as its significance in improving clinical symptoms and overall prognosis. However, due to the lack of clinical research evidence on the relationship between nutrition and AD-related cognitive impairment. This study will start from the influencing factors of malnutrition in elderly MCI patients and explore the prediction model of malnutrition in elderly patients with mild cognitive impairment to make up for the lack of evidence in clinical studies on the relationship between nutrition and AD-related cognitive impairment.

[0032] This review examines previous studies on malnutrition in elderly patients with mild cognitive impairment and identifies potential influencing factors. These include, but are not limited to, biochemical markers such as potassium (K), sodium (Na), and cholesterol, as discussed in this article, as well as sociodemographic and lifestyle factors such as activities of daily living (ADL), education level, and residential status. Although some studies have reported risk factors for malnutrition in elderly patients with mild cognitive impairment, subsequent studies have not fully addressed these risk factors and have relied solely on traditional risk prediction models. While these models have some predictive efficacy, they are clearly inadequate in assisting decision-making and providing targeted recommendations.

[0033] Decision tree models offer unique advantages. They can process predictive variables without creating dummy variables, and can intuitively present various risk factors through a tree-like structure, making them easier for medical staff to identify and understand. Given this, the present invention selected a decision tree algorithm to deeply analyze the intrinsic relationships and hierarchical classification characteristics between various factors in patients with MCI and malnutrition. The classification and regression tree (CART) method, a dichotomous statistical technique, starts with the entire sample and gradually divides it into subsamples with high homogeneity relative to predefined results, thereby accurately identifying the variables that most significantly influence the predicted results. Through automatic analysis of the optimal threshold, the input variables that contribute to the most effective division are dichotomized.

[0034] The indicators incorporated in this paper are more comprehensive and systematic, encompassing not only potential risk factors for cognitive function in the elderly, but also traditional risk factors and laboratory indicators such as BMI, number of chronic comorbidities, oral health, appetite, potassium, sodium, and cholesterol. Based on this, the present invention innovatively combines a decision tree model with a logistic regression model to construct an accurate prediction model for malnutrition risk in elderly patients with MCI, aiming to provide a practical reference for clinical prevention and treatment.

[0035] Specifically, the method of predicting malnutrition in elderly patients with mild cognitive impairment is analyzed as follows: This paper adopts the convenience sampling method to select data from two tertiary hospitals in two corresponding regions for cross-sectional analysis, which specifically includes: Step 1: Determine the analysis object for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment; In a preferred but non-limiting embodiment of the present invention, in step 1, a cross-sectional survey is conducted on elderly people diagnosed with mild cognitive impairment (MCI) in the corresponding hospital. The present invention includes subjects recruited between October 2023 and November 2024. The inclusion criteria for analysis and prediction of malnutrition in elderly patients with mild cognitive impairment are as follows: (3) Age ≥ 60 years; (4) diagnosed with MCI; (3) self-reported no visual or hearing impairment; (4) Able to sign informed consent.

[0036] In a preferred but non-limiting embodiment of the present invention, in step 1, the screening criteria for confirming MCI include: i. The subject or informant reported a relative decline in cognitive function in the past year; ii. The subject has objective cognitive impairment that is inconsistent with age and education level; iii. The ability to carry out activities of daily living is intact (Lawton-Brody Activities of Daily Living score ≤ 16 points); iv. No dementia; Subjects were excluded from the diagnosis of MCI if they had the following conditions: (1) A history of serious neurological, mental or other diseases that may affect brain function; (2) taking any medication that may impair or improve cognitive function (such as memantine and antipsychotic drugs) in the past 6 months; (3) Participated in another cognitive function-related training within 1 year. Subjects are allowed or required to withdraw from the trial if the following circumstances occur: a serious medical accident occurs during the education or follow-up period; or the subject is unwilling to continue participating in the trial.

[0037] In a preferred but non-limiting embodiment of the present invention, in step 1, the subject has objective cognitive impairment that is inconsistent with age and education level, including: Those with an elementary school education had a score of less than 18 on the Montreal Cognitive Assessment-Basic (MoCA-B), those with a secondary school education had a score of less than 21, and those with a higher education had a score of less than 23.

[0038] In a preferred but non-limiting embodiment of the present invention, in step 1, the ability to perform daily living activities is intact means that the subject's Lawton-Brody score for daily living activities is ≤16 points.

[0039] Step 2: Perform group nutritional screening on the subjects to be analyzed to predict malnutrition in elderly patients with mild cognitive impairment; In a preferred but non-limiting embodiment of the present invention, in step 2, the present invention uses the 2002 Nutritional Risk Screening Tool (NRS-2002) to perform group nutritional screening on subjects analyzed for predicted malnutrition in elderly patients with mild cognitive impairment. This tool was developed by the European Society for Clinical Nutrition and Metabolism (ESPEN) in 2002 to simplify the nutritional risk screening process for hospitalized patients and improve the efficiency of clinical nutritional assessments.

[0040] In a preferred but non-restrictive embodiment of the present invention, in step 2, the NRS-2002 scale of the 2002 version of the nutritional risk screening tool is assessed from three dimensions: impaired nutritional status, disease severity and age, and the sum of the scores of each dimension constitutes the screening total score, i.e., an NRS-2002 total score lower than 3 points indicates a good nutritional status, while 3 points and above indicate the presence of malnutrition. Since its launch, NRS-2002 has been widely recognized for its scientificity and practicality, and the Chinese Medical Association's Parenteral and Enteral Nutrition Branch and the Chinese Medical Association recommend it for nutritional risk assessment. In the present invention, an NRS-2002 score of 3 points and above is determined to be at risk of malnutrition, while a score lower than 3 points indicates no risk of malnutrition.

[0041] Step 3: Perform malnutrition risk prediction for the analysis object of analyzing and predicting malnutrition in elderly patients with mild cognitive impairment.

[0042] In a preferred but non-limiting embodiment of the present invention, step 3 specifically comprises: Step 3-1: Identify demographic characteristics and lifestyle factors for malnutrition risk prediction; In a preferred but non-limiting embodiment of the present invention, in step 3-1, demographic characteristics and lifestyle factors are retrospectively collected by professionally qualified investigators through face-to-face interviews. Demographic characteristics include the subject's age, gender, education level, marital status, household income, residential address, medical insurance, polypharmacy, and number of chronic comorbidities for predicting malnutrition in elderly patients with mild cognitive impairment.

[0043] In a preferred but non-limiting embodiment of the present invention, in step 3-1, lifestyle factors include analyzing and predicting information such as appetite, daily eating habits, regular physical activity, sleep quality, self-care ability, caregiver efficacy, and dietary diversity of the analysis subject for malnutrition in elderly patients with mild cognitive impairment; Appetite was assessed using the Simplified Nutrition Appetite Questionnaire (SNAQ), which includes four parts: appetite, satiety during meals, food taste, and number of meals per day, with a total score ranging from 4 to 20 points; Daily eating habits and regular physical activity were binary variables, defined as follows: daily eating habits refer to whether or not eating regularly; regular eating refers to regular meal times and frequencies, and eating at least three meals a day on time is considered regular eating; regular physical activity refers to walking, square dancing, cycling, and other activities for 150 minutes or more per week in the past month; Sleep quality was assessed using the Pittsburgh Sleep Quality Index (PSQI), a scale developed by Buysse et al. It assesses patients' sleep status over the past month and covers seven dimensions: subjective sleep quality, sleep onset time, sleep duration, sleep efficiency, sleep disturbances, hypnotic use, and daytime dysfunction. It consists of 19 self-rated items and 5 peer-rated items (not included in the total score). Each item is scored on a scale of 0-3, with higher scores indicating poorer sleep quality. The Chinese version of the PSQI has a Cronbach's α coefficient of 0.82-0.83 and has been used in cross-sectional studies of sleep quality in patients with dementia and in cohort studies of cognitive decline and sleep quality in healthy elderly people and patients with mild cognitive impairment.

[0044] Self-care ability was measured using the Activities of Daily Living (ADL) scale, which assesses basic and instrumental activities of daily living. The scale consists of 20 items, each scored 1 to 4 (1: completely independent; 2: with some difficulty but still possible; 3: requiring assistance; 4: completely unable to do). The total score ranges from 20 to 80. A score of 20 indicates complete independence, while a score greater than 20 indicates varying degrees of decline in ADL.

[0045] Caregiver efficacy was measured using the General Self-Efficacy Scale (GSES) for primary caregivers. This scale was developed by German psychologist Schwarzer et al. in 2001. Wang Caikang et al. completed reliability and validity testing of the Chinese version of the GSES, with good results. It is suitable for measuring self-efficacy in non-specific areas. The scale consists of 10 dimensions and is scored on a 4-point Likert scale, with "completely true" (4 points), "mostly true" (3 points), "somewhat true" (2 points), and "completely false" (1 point). The total score ranges from 10 to 40, with higher scores indicating greater self-efficacy.

[0046] Dietary diversity is calculated using the Dietary Diversity Score (DDS). Based on the "Chinese Food Composition Table 2002" and the "Balanced Diet Pagoda for Chinese Residents," food is categorized into nine categories: cereals, vegetables, fruits, legumes, eggs, meat, fish, dairy products, and oils. The DDS is calculated by summing the types of food consumed over three days, with each food category receiving one point, for a total of nine points.

[0047] Step 3-2: Identify physiological indicators for malnutrition risk prediction; In a preferred but non-limiting embodiment of the present invention, in step 3-2, the physiological indicators include analyzing and predicting the malnutrition of the elderly patients with mild cognitive impairment, the body mass index (BMI), oral health status, overall cognitive function, disorientation, immediate memory impairment, attention deficit, retrograde amnesia, language disorder, neuropsychiatric symptoms, dysphagia, hemoglobin, blood potassium, blood sodium, cholesterol and albumin of the analysis subject; BMI is an internationally recognized indicator for measuring body fat content and health status. The calculation formula is: BMI = weight (kg) / height² (m²); Oral health status was assessed using the Oral Health Assessment Tool (OHAT), which is widely used for oral health screening of the elderly. It includes eight items: lips, tongue, gums and tissues, saliva, natural teeth, dentures, oral hygiene, and toothache. Each item is scored 0-2 points (0 point: healthy state; 1 point: changed health state; 2 points: unhealthy state). The lower the total score, the better the oral health status. Global cognitive function was assessed using the Mini-Mental State Examination (MMSE), developed by Folstein et al. in 1975. The MMSE covers items such as orientation to time and place, immediate memory and recall, attention and calculation, and language. The maximum score is 30, with higher scores indicating better cognitive function. Specifically, for individuals with a secondary school education or higher, a score >24 indicates normal cognitive function; for individuals with only an elementary school education, a score >20 indicates normal cognitive function; for individuals with no education, a score >17 indicates normal cognitive function; otherwise, cognitive impairment is diagnosed. The Chinese version of the MMSE, developed by Wang Zhengyu et al. in 1989, has a test-retest reliability of 0.91. Disorientation, immediate memory impairment, attention deficit, retrograde amnesia, and language impairment are also measured using this scale.

[0048] Neuropsychiatric symptoms were measured using the Neuropsychiatric Inventory (NPI), developed by Cummings et al. in 1994. The Chinese version of the NPI-Q was translated by Ma Wanxin et al. in 2010 and completed by informants. It demonstrated good reliability and validity when administered to 51 patients with dementia and their informants.

[0049] Dysphagia is assessed using the Kubota drinking test. Patients sit upright and drink 30 ml of warm water. The drinking process and duration are observed. The drinking process is categorized as follows: ① Complete in one gulp without choking; ② Complete in two or more gulps without choking; ③ Complete in one gulp with choking; ④ Complete in two or more gulps with choking; ⑤ Multiple choking and inability to complete the drink. The criteria for diagnosing dysphagia are: if the drinking process is ① and takes less than 5 seconds, swallowing function is normal; if the drinking process is ① but takes 5 seconds or more, or if the drinking process is ②, swallowing function is suspected to be abnormal; if the drinking process is ③-⑤, swallowing function is abnormal.

[0050] Biochemical indices such as serum potassium, serum sodium, cholesterol, albumin, and hemoglobin were collected through the hospital electronic medical record system.

[0051] Step 3-3: Perform statistical analysis and prediction on physiological indicators, demographic characteristics and lifestyle factors.

[0052] In a preferred but non-limiting embodiment of the present invention, in step 3-3, the nutritional status, age, gender, education level, marital status, family income, residential address, medical insurance, multiple medication status, number of chronic disease comorbidities, appetite, daily eating habits, regular physical activity, sleep status, self-care ability, caregiver efficacy, dietary diversity, BMI, oral health status, overall cognitive function, disorientation, immediate memory impairment, attention deficit, retrograde amnesia, language disorder, neuropsychiatric symptoms, dysphagia, hemoglobin, blood potassium, blood sodium, cholesterol and albumin of all analysis subjects are collected. These data can be divided into a "malnutrition risk group" and a "normal nutrition group" according to self-defined criteria; The test data were statistically analyzed using R-4.4.2 and SPSS27.0 software. If the test data followed a normal distribution, they were expressed as mean ± standard deviation (x ± s), and the independent sample t-test was used. If the test data followed a non-normal distribution, they were expressed as the median (quartile) of the test data (M(Qn)), and the median of the test data was tested using the Mann-Whitney U test. If the test data were expressed as the number of cases (percentage) (n(%)), the chi-squared test was used. Lasso regression was used to reduce the dimensionality of the monitoring data that passed the test as included variables to screen out the influencing factors of malnutrition in elderly patients with mild cognitive impairment. In addition, binary logistic regression and decision tree were used to establish a clinical prediction model for the monitoring data as an evaluation model. The Hosmer-Lemeshow test was used to evaluate the goodness of fit of the model and to draw a calibration curve. The area under the receiver operating characteristic curve (AUC) was used to determine the discrimination of the evaluation model.

[0053] Regarding the present method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment, a total of 176 questionnaires were distributed, of which 170 were valid, for a response rate of 96.6%. Among the respondents, 66 patients (38.82%) were in the malnutrition risk group, and 104 patients (61.18%) were in the normal nutrition group.

[0054] With malnutrition risk as the dependent variable (no = 0, yes = 1), the following 11 factors were statistically significant after univariate analysis and were subsequently included as independent variables in the lasso regression model: age, number of comorbid chronic diseases, appetite, self-care ability, caregiver efficacy, body mass index (BMI), oral health status, dysphagia, hemoglobin, cholesterol, and albumin.

[0055] Lasso regression was used to reduce the dimensionality of the included variables. Correlations between variables were analyzed using ten-fold cross-validation, and the Lasso coefficient curve for the included variables was obtained. The model was optimal when the distance from the mean square error (λ1se) was one standard error (λ1se = 0.03749). Seven predictors with non-zero coefficients were identified as factors influencing malnutrition in elderly patients with mild cognitive impairment: BMI, number of comorbid chronic diseases, oral health, appetite, hemoglobin, dysphagia, and self-care ability.

[0056] Seven factors that were statistically significant in both univariate and lasso regression analyses—BMI, number of chronic disease comorbidities, oral health status, appetite, hemoglobin, dysphagia, and self-care ability—were included as independent variables in a binary logistic regression model. Binary logistic regression analysis showed that the following variables were risk factors for cognitive impairment in the elderly: BMI, number of chronic disease comorbidities, and oral health status (P < 0.05). The results are shown in Table 1. Binary logistic regression analysis of risk factors for malnutrition in elderly patients with mild cognitive impairment: Table 1

[0057] The classification and regression tree (CART) method was used, which is a dichotomous statistical technique. The significance level of the decision tree growth branch split was set to 0.05. The minimum sample size of the parent node was set to 10, and the minimum sample size of the child node was set to 5. If the sample size of a node did not meet these requirements, the node was considered a terminal node and no further splitting was performed. Variables that were statistically significant in the binary logistic regression analysis were included. The present invention constructed a decision tree with 4 layers and 11 terminal nodes, and screened out 5 explanatory variables: body mass index (BMI), oral health status (OHAT), number of chronic disease comorbidities, hemoglobin, and appetite (SNAQ), as shown in Figure 5. Figure 3 shown.

[0058] The root node divides the total sample (n=170) into two groups based on BMI. Among the samples with BMI≤20.695 (n=30), the proportion of samples with a lower risk of malnutrition is relatively high (86.7%), and this branch is further subdivided according to oral health status. Among the samples with BMI>20.695 (n=140), the proportion of samples with a higher risk of malnutrition is 71.4%, and this branch is divided according to variables such as the number of chronic disease comorbidities and appetite. Specifically, when the number of chronic disease comorbidities is less than or equal to 2 (n=93), the proportion of samples with a higher risk of malnutrition is 83.9%. When the number of chronic disease comorbidities is >2 (n=47), different malnutrition risk distribution characteristics are presented according to the segmentation based on hemoglobin and appetite. Figure 3 As shown in the decision tree, when the hemoglobin value is less than or equal to 133.500, the risk of malnutrition is relatively high. For example, in node 7, the proportion of malnutrition cases reaches 33.3%. When the hemoglobin value is greater than 133.500, the risk of malnutrition is relatively low, as shown in node 8, where the proportion of malnutrition cases is only 21.4%. Appetite is also an important factor. When the appetite score is less than or equal to 14.5, the risk of malnutrition is high. For example, in node 9, the proportion of malnutrition cases is as high as 8.3%. When the appetite score is greater than 14.5, the risk of malnutrition is relatively low, as shown in node 10, where the proportion of malnutrition cases is 47.6%.

[0059] like Figure 4 As shown in Table 2, the area under the receiver operating characteristic curve (AUC) for the logistic regression model was 0.819 (95% confidence interval: 0.752 - 0.887). The AUC for the decision tree model was 0.824 (95% confidence interval: 0.753 - 0.895). As shown in Table 2, comparing the classification performance of the logistic regression model and the decision tree model, the decision tree model slightly outperformed the logistic regression model.

[0060] Table 2

[0061] This study constructed a decision tree model that comprehensively considers demographic characteristics, lifestyle factors, and other information to help identify malnutrition risk in elderly patients with mild cognitive impairment (MCI). This is the first study to utilize classification and regression tree (CART) decision tree analysis to identify malnutrition risk in elderly patients with MCI. The key finding of this study is that the constructed decision tree model exhibits comparable predictive value and good internal validation compared to a binary logistic regression model. The decision tree uses an intuitive tree-like structure to present the distribution of malnutrition risk under different variable combinations, capturing interactions and nonlinear relationships between variables. Extensive research and in-depth validation have confirmed that body mass index (BMI), the first splitting variable, plays the most critical role among all considered variables in distinguishing patients at risk of malnutrition, being the primary consideration. Subsequently, oral health status, number of comorbid chronic diseases, hemoglobin, and appetite variables appear in descending order of importance.

[0062] BMI and oral health status are important potentially modifiable risk factors for malnutrition in elderly patients with MCI. The results of this study are generally consistent with previous studies. Data analysis revealed significant differences in BMI and oral health status (OHAT) between high- and low-risk groups for malnutrition. Specifically, a logistic regression model showed a regression coefficient (β) of -0.198 for BMI, a P value of 0.006, and an odds ratio (OR) of 0.82, indicating a negative correlation between BMI and malnutrition risk. This suggests that a low BMI is often associated with an increased risk of malnutrition. This may be because low BMI is often associated with insufficient nutritional reserves, which in turn affects physical function and leads to an increased risk of malnutrition. Furthermore, the regression coefficient (β) for OHAT was 0.3, a P value of 0.003, and an OR of 1.349, indicating that poor oral health (higher OHAT scores) is associated with an increased risk of malnutrition. This may be because oral health problems can lead to difficulty eating and decreased appetite. Over time, these problems can lead to nutritional deficiencies, increasing the risk of malnutrition. For example, patients with poor oral health may be forced to change their eating habits due to difficulty in eating, and tend to choose foods that are easier to chew and swallow but may be nutritionally unbalanced. This change can easily lead to nutritional imbalance in the body and further increase the risk of malnutrition.

[0063] A decision tree model further confirmed these results in a more intuitive and interpretable manner. The decision tree first split based on BMI. In the branch with a BMI ≤ 20.695, the risk of malnutrition was relatively low (e.g., 13.3% risk in node 1); whereas, in the branch with a BMI > 20.695, the risk of malnutrition was relatively high (e.g., 71.4% risk in node 2), consistent with the finding in logistic regression that low BMI was associated with a higher risk of malnutrition. Furthermore, within the branch with a BMI ≤ 20.695, further splitting based on OHAT revealed a significant difference in the proportion of malnutrition risk between those with OHAT ≤ 0.5 (node ​​3) and those with OHAT > 0.5 (node ​​4), again demonstrating the important impact of oral health on malnutrition risk, consistent with the positive correlation between OHAT and malnutrition risk in logistic regression.

[0064] The logistic regression results of the present invention also showed that elderly MCI patients with poor oral health were at higher risk of malnutrition than those with good oral health. Poor oral health can lead to a range of problems, such as difficulty eating and decreased appetite, which negatively impacts nutritional intake and ultimately increases the risk of malnutrition. However, improving oral health may improve patients' ability to eat and appetite, thereby reducing the risk of malnutrition. While improving oral health may have a positive effect on reducing the risk of malnutrition, individual factors such as underlying medical conditions, age, and psychological status also have significant impacts on nutritional status and need to be comprehensively considered in research and clinical practice. Therefore, in addition to focusing on oral health as a key factor, comprehensive management measures including a balanced diet, appropriate nutritional supplements, and psychological support are crucial for reducing the risk of malnutrition in elderly MCI patients. Early nutritional screening and better nutritional care can reduce the incidence of malnutrition. The European Society for Clinical Nutrition and Metabolism (ESPEN) guidelines recommend that anyone over 65 years of age should undergo nutritional risk screening, especially those who are frail, have chronic non-communicable diseases (such as diabetes and cancer), live alone, rely on social services or require assistance with daily living, and are hospitalized. Nutritional risk screening should be performed on outpatients and inpatients in the cognitive impairment department during their first visit.

[0065] Previous studies have shown an association between malnutrition risk and cognitive decline. However, this association was not detected in the present invention. One possible reason for this difference is that the present invention used convenience sampling, which may have led to selection bias. At the same time, previous studies found that the risk of malnutrition in elderly MCI patients was related to diet, cognition, living arrangements, income, insurance, sleep and exercise, but the present invention did not confirm this. Previous studies compared MCI patients with normal elderly people, emphasizing nutritional changes at different stages of cognitive impairment; while the present invention focused on the differences within MCI patients, focusing on analyzing the causes of their internal variations and factors related to malnutrition risk. Different research purposes led to different results.

[0066] The decision tree structure is similar to a folded graph and can extract classification rules from irregular situations. It determines the branches under the node by comparing the attribute values ​​of each internal node and draws classification conclusions for the leaf nodes. To make the results more reliable and robust, the present invention further performs decision tree analysis based on the CART algorithm. In this invention, the decision tree model determines in an intuitive and interpretable way that BMI, oral health status, number of chronic disease comorbidities, hemoglobin, and appetite are important risk factors for malnutrition in elderly patients with MCI.

[0067] BMI is the most important risk factor for malnutrition in elderly patients with MCI. A BMI of 20.695 or less increases the risk of malnutrition. BMI is often used to measure body fatness and health status, and abnormal BMI is closely associated with MCI. A decrease in BMI often precedes the onset of clinical Alzheimer's disease (AD) symptoms and is considered a marker of cognitive decline or MCI. The underlying mechanism is unclear, but it may be because the pathological changes predate the clinical onset of AD. The relationship between BMI and MCI is not uniform. Those whose BMI initially increases in early middle age and then decreases in late middle age are at higher risk of dementia compared to those who do not. In fact, malnutrition is not solely related to BMI but is a complex issue influenced by multiple factors. This may be because a lower BMI often reflects insufficient nutritional reserves. Furthermore, as MCI progresses, decreased physical function may affect nutrient absorption and utilization, potentially contributing to the higher incidence of malnutrition in those with a lower BMI.

[0068] However, this does not mean that elderly MCI patients with a higher BMI are not at risk for malnutrition. Elderly MCI patients with numerous comorbid chronic diseases may experience altered metabolism and nutritional requirements, a common risk factor for malnutrition in the elderly. Comorbidities are common among elderly patients, and the coexistence of multiple diseases can impact nutritional status. Therefore, malnutrition in the elderly is very common and can occur simultaneously or independently. The long-term inflammatory stimulation of chronic diseases can reduce appetite and lead to a negative nitrogen balance, leading to adverse consequences such as frailty and malnutrition. The more chronic diseases elderly patients have, the higher their risk of malnutrition. This may be due to the long-term coexistence of multiple chronic diseases in the elderly, which can cause chronic damage to organ function and lead to imbalanced energy metabolism, triggering a long-term vicious cycle of metabolic imbalance and prolonged disease course. Long-term medication use in elderly patients with chronic diseases can reduce the body's intake and absorption of nutrients. Elderly patients with fewer chronic diseases tend to have a stronger awareness of disease prevention, are more determined and committed to health management, and place greater emphasis on a nutritious and healthy lifestyle in terms of diet and lifestyle.

[0069] The results of the present invention found that poor oral health has a negative impact on the nutritional status of elderly patients with MCI, which is consistent with the results of related research. Some studies have shown that the incidence of oral frailty in elderly Chinese people ranges from 8.1% to 53.2%. Compared with healthy people of the same age, patients with cognitive impairment have worse oral health and are more likely to develop problems such as caries, residual roots, periodontitis, and oral mucosal inflammation. Oral frailty in the elderly plays an important role in predicting various adverse health outcomes associated with cognitive impairment, such as malnutrition. If timely detection and intervention can be carried out, it can help avoid or delay the occurrence and development of adverse events.

[0070] A hemoglobin level below 133.5 increases the incidence of malnutrition in elderly patients with MCI. Hemoglobin is often used as a basis for clinical nutritional assessments. Due to decreased appetite, swallowing discomfort, and impaired absorption and digestion, insufficient food intake and a lack of raw materials for hemoglobin synthesis are likely to occur, leading to nutritional iron-deficiency anemia. Therefore, medical staff should pay attention to patients' hemoglobin and albumin levels and provide timely nutritional fortification and supplementation when necessary. A vicious cycle may exist between malnutrition and hemoglobin. Hemoglobin transports oxygen and carbon dioxide. A decrease in hemoglobin directly leads to a decrease in the body's oxygen-carrying capacity, anemia, and malnutrition. Hemoglobin is the main component of red blood cells. When the body lacks protein, the hemoglobin content decreases. Malnutrition in the body exacerbates protein depletion, leading to a decrease in hemoglobin.

[0071] Loss of appetite is a common behavioral and psychiatric symptom in patients with mild cognitive impairment (MCI), which significantly reduces their food intake and is a possible factor leading to malnutrition. The conclusions of this invention are highly consistent with the results of previous studies. Elderly patients are prone to malnutrition due to digestive dysfunction, reduced food intake, and disease status, which increases the risk of adverse clinical outcomes and death. In the early stages of cognitive decline, even years before the onset of the disease, pathological changes in the olfactory system are believed to be the cause of reduced nutrient intake and weight loss. Olfactory dysfunction is associated with reduced diet quality, diversity, and energy density. Elderly MCI patients may have their food intake affected by decreased sense of smell and taste, as well as changes in appetite.

[0072] Based on the results of this study, it is recommended that managers focus on monitoring BMI and oral health in the elderly to prevent malnutrition in elderly MCI patients. Monitoring mechanisms for BMI and oral health in elderly MCI patients should be gradually improved, and guidance and intervention on a balanced diet for these patients should be strengthened to reduce the incidence of malnutrition in these patients.

[0073] The present invention is only a cross-sectional study and therefore cannot provide a causal explanation. In-depth longitudinal studies should be conducted in the future to further understand the trajectory of factors affecting cognitive impairment in the elderly. In addition, the sample size of the present invention is relatively limited, which may affect the extrapolation and generalizability of the research results to a certain extent. Despite these limitations, the present invention still has three key advantages. First, the variables were selected through a systematic literature review to ensure the comprehensiveness and systematicness of the indicators. Secondly, lasso regression analysis was applied for strict variable screening, which effectively avoided variable collinearity, ensured the independence and validity of the variables included in the model, and facilitated subsequent analysis. Third, a decision tree model was introduced. The model is simple, clear, intuitive and well-structured, presenting data and variable relationships in an easy-to-understand way, enabling healthcare professionals to make risk-based decisions more efficiently and accurately, and providing support for clinical practice and health management.

[0074] This study is the first to conduct a decision tree analysis of risk factors for malnutrition in elderly patients with MCI. Despite certain limitations, this analysis provides valuable insights. Compared with other risk factors for malnutrition, factors such as BMI, number of comorbid chronic diseases, hemoglobin levels, and appetite significantly increase the risk of malnutrition in elderly patients with MCI.

[0075] Furthermore, the present invention employs a decision tree model for analysis. As a simple, intuitive, and practical hierarchical approach, decision trees can help healthcare professionals make more effective risk-based decisions. Therefore, promoting this model in future medical research is worthwhile. Future research is warranted to expand the sample size and continue to explore new indicators of malnutrition in elderly patients with MCI.

[0076] like Figure 2 As shown, the system for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to the present invention comprises: Determine module 1, which is used to determine the analysis object for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment; Determine a second module for performing group nutrition screening on analysis subjects for predicting malnutrition in elderly patients with mild cognitive impairment; The prediction module is used to perform malnutrition risk prediction for an analysis object that analyzes and predicts malnutrition in elderly patients with mild cognitive impairment.

[0077] The beneficial effects of the present invention are that, compared with the prior art, the technical effects of the present invention include: This study incorporates a more comprehensive and systematic indicator system, encompassing potential risk factors, traditional risk factors, and laboratory data for elderly patients with mild cognitive impairment. This represents a comprehensive breakthrough in research perspectives and overcomes the limitations of existing technologies. By comprehensively comparing decision tree models with logistic regression models, a precise malnutrition risk prediction model for elderly patients with MCI has been constructed. This provides a valuable reference for clinical prevention and treatment, and is expected to promote the optimization and improvement of malnutrition prevention and treatment strategies for elderly patients with mild cognitive impairment.

[0078] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that the specific implementation methods of the present invention can still be modified or replaced with equivalents, and any modifications or equivalent replacements that do not deviate from the spirit and scope of the present invention should be covered within the protection space of the claims of the present invention.

Claims

1. A method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment, characterized in that: include: Convenience sampling was used to select data from corresponding hospitals for cross-sectional analysis, which specifically included: Step 1: Determine the analysis object for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment; Step 2: Perform group nutritional screening on the subjects to be analyzed to predict malnutrition in elderly patients with mild cognitive impairment; Step 3: Perform malnutrition risk prediction for the analysis object of analyzing and predicting malnutrition in elderly patients with mild cognitive impairment.

2. The method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to claim 1, characterized in that: In step 1, the inclusion criteria for the analysis of the prediction of malnutrition in elderly patients with mild cognitive impairment were determined as follows: Age ≥60 years; Diagnosed with MCI; (3) self-reported no visual or hearing impairment; (4) Able to sign informed consent.

3. The method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to claim 2, characterized in that: In step 1, the operational criteria used for screening to confirm MCI include: i. The subject or informant reported a relative decline in cognitive function in the past year; ii. The subject has objective cognitive impairment that is inconsistent with age and education level; iii. The ability to carry out activities of daily living is intact; iv. No dementia; Subjects were excluded from the diagnosis of MCI if they had the following conditions: (1) A history of serious neurological, mental or other diseases that may affect brain function; (2) Taking any medication that may impair or improve cognitive function in the past 6 months; (3) Participated in another training related to cognitive function within 1 year.

4. The method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to claim 3, characterized in that: In step 1, the subject has objective cognitive impairment that is inconsistent with age and education level, including: Those with primary education have a score of less than 18 on the Montreal Cognitive Assessment Basic version; those with secondary education have a score of less than 21; and those with higher education have a score of less than 23; In step 1, the ability to perform daily living activities is intact if the subject's Lawton-Brody score for daily living activities is ≤16 points.

5. The method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to claim 4, characterized in that: In step 2, the 2002 version of the Nutritional Risk Screening Tool was used to perform group nutritional screening on the subjects who were analyzed to predict malnutrition in elderly patients with mild cognitive impairment; In step 2, the NRS-2002 scale of the 2002 version of the nutritional risk screening tool is used to assess the nutritional status, disease severity and age from three dimensions. The sum of the scores of each dimension constitutes the total screening score. That is, an NRS-2002 total score below 3 points indicates good nutritional status, while 3 points or above indicates malnutrition.

6. The method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to claim 5, characterized in that: Step 3 specifically includes: Step 3-1: Identify demographic characteristics and lifestyle factors for malnutrition risk prediction; Step 3-2: Identify physiological indicators for malnutrition risk prediction; Step 3-3: Perform statistical analysis and prediction on physiological indicators, demographic characteristics and lifestyle factors.

7. The method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to claim 6, characterized in that: In step 3-1, demographic characteristics include age, gender, education level, marital status, family income, residential address, medical insurance, polypharmacy, and number of chronic diseases in the subjects for predicting malnutrition in elderly patients with mild cognitive impairment; In step 3-1, lifestyle factors include analyzing the appetite, daily eating habits, regular physical activity, sleep quality, self-care ability, caregiver efficacy, and dietary diversity information of the subjects to predict malnutrition in elderly patients with mild cognitive impairment; Appetite was assessed using the simplified nutritional appetite questionnaire; Daily eating habits and regular physical activity were dichotomous variables, defined as follows: daily eating habits refer to whether or not eating regularly; regular eating refers to regular meal times and frequencies, and eating at least three meals a day on time is considered regular eating; regular physical activity refers to walking, square dancing, or cycling for 150 minutes or more per week in the past month; Sleep quality was assessed using the Pittsburgh Sleep Quality Index; Self-care ability was measured using the Activities of Daily Living Scale, which assesses basic activities of daily living and instrumental activities of daily living; Caregiver efficacy was measured using the General Self-Efficacy Scale for Primary Caregivers; Dietary diversity was calculated based on the dietary diversity score.

8. The method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to claim 7, characterized in that: In step 3-2, the physiological indicators include analyzing the body mass index, oral health status, global cognitive function, disorientation, immediate memory impairment, attention deficit, retrograde amnesia, language disorder, neuropsychiatric symptoms, dysphagia, hemoglobin, blood potassium, blood sodium, cholesterol and albumin of the analyzed subjects for predicting malnutrition in elderly patients with mild cognitive impairment; BMI is an internationally recognized indicator for measuring body fat content and health status. The calculation formula is: BMI = weight (kg) / height² (m²); Oral health status was assessed using an oral health assessment tool; Global cognitive function was assessed using the Mini-Mental State Examination; Neuropsychiatric symptoms were measured using the Neuropsychiatric Rating Scale; Dysphagia was assessed using the Kubota water drinking test to assess the patient's swallowing function; Biochemical indices such as serum potassium, serum sodium, cholesterol, albumin, and hemoglobin were collected through the hospital electronic medical record system.

9. The method for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment according to claim 8, characterized in that: In step 3-3, all subjects were collected for nutritional status, age, sex, education level, marital status, family income, residential address, medical insurance, polypharmacy, number of chronic diseases, appetite, daily eating habits, regular physical activity, sleep status, self-care ability, caregiver efficacy, dietary diversity, BMI, oral health status, global cognitive function, disorientation, immediate memory impairment, attention deficit, retrograde amnesia, language disorder, neuropsychiatric symptoms, dysphagia, hemoglobin, blood potassium, blood sodium, cholesterol, and albumin. The test data were statistically analyzed using R-4.4.2 and SPSS27.0 software. If the test data followed a normal distribution, they were expressed as mean ± standard deviation (x ± s), and an independent sample t-test was used. If the test data followed a non-normal distribution, they were expressed as the median of the test data, and the Mann-Whitney U test was used for the median of the test data. If the test data were expressed as the number of cases, a chi-squared test was used. Lasso regression was used to reduce the dimensionality of the monitoring data that passed the test and identify the influencing factors of malnutrition in elderly patients with mild cognitive impairment. In addition, binary logistic regression and decision tree were used to establish a clinical prediction model for the monitoring data as an evaluation model. The Hosmer-Lemeshow test was used to evaluate the goodness of fit of the model and to draw a calibration curve. The area under the receiver operating characteristic curve was used to determine the discrimination of the evaluation model.

10. A system for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment, characterized by: include: Determine module 1, which is used to determine the analysis object for analyzing and predicting malnutrition in elderly patients with mild cognitive impairment; Determine a second module for performing group nutrition screening on analysis subjects for predicting malnutrition in elderly patients with mild cognitive impairment; The prediction module is used to perform malnutrition risk prediction for an analysis object that analyzes and predicts malnutrition in elderly patients with mild cognitive impairment.

Images