Prediction method and system for anxiety mood morbidity rate of diabetic child patient

Abstract

The invention relates to a method and system for predicting the anxiety mood prevalence rate of a diabetic child patient, and belongs to the field of artificial intelligence prediction. According to the method, multi-dimensional clinical data such as demographic data, treatment modes, management behaviors and metabolic indexes of diabetic children are collected, evaluation results of a children anxiety mood disorder screening table (SCARED) are combined, single-factor analysis, correlation analysis and multi-factor Logistic regression analysis are carried out, and an anxiety mood prediction model is constructed. The model takes age, an insulin treatment mode and glycosylated hemoglobin as core variables, and can accurately predict the anxiety risk of a child patient. The system comprises a data acquisition module, a psychological assessment module, an analysis modeling module, a risk prediction module and a report generation module, and realizes automatic data integration, model construction and individual intervention suggestion output. Clinical verification shows that the model prediction efficiency is excellent, and the method has good clinical application value and is beneficial to early recognition and intervention of anxiety mood of diabetic children and improvement of physical and psychological health and disease management quality of the diabetic children.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence prediction, and in particular to a method and system for predicting the prevalence of anxiety in children with diabetes. Background Technology

[0002] Type 1 diabetes mellitus (T1DM) is a chronic autoimmune disease caused by a combination of genetic and environmental factors, characterized by the destruction of pancreatic beta cells and an absolute deficiency of insulin secretion. Globally, the incidence of T1DM is increasing at a rate of approximately 1.8% per year. According to the International Diabetes Federation (IDF), there are currently more than 1.2 million children and adolescents under the age of 20 with this disease. Although the incidence rate in China is lower than in Europe and the United States, the annual growth rate is as high as 4.1%, showing a rapid upward trend and posing a serious challenge to public health systems both globally and in China.

[0003] The management of type 1 diabetes mellitus (T1DM) requires children and their families to strictly adhere to daily insulin therapy, blood glucose monitoring, dietary control, and regular exercise. This complex and lifelong treatment burden easily triggers a series of psychological problems, with anxiety and depression being the most common. Studies have shown that the prevalence of anxiety in children with T1DM is significantly higher than in healthy peers. This not only severely impairs the quality of life of children with T1DM, but also leads to worsening of blood glucose control by affecting treatment adherence (such as missing insulin injections and avoiding blood glucose monitoring) and activating the hypothalamic-pituitary-adrenal (HPA) axis, which increases the secretion of stress hormones. This creates a vicious cycle of "poor blood glucose control - increased anxiety - further loss of blood glucose control," ultimately increasing the risk of diabetic ketoacidosis (DKA) and long-term microvascular complications.

[0004] In recent years, although the academic community has gradually recognized the importance of psychological problems in children with type 1 diabetes mellitus (T1DM), the level of attention, recognition, and intervention in clinical practice remains low. Current research mainly focuses on describing the prevalence of anxiety and depression or their correlation with single indicators (such as glycated hemoglobin HbA1c), lacking a systematic exploration of the synergistic effects of multiple factors. The generation of anxiety is a complex result of the synergistic effects of physiological, therapeutic, and socio-psychological factors. For example, as an advanced treatment method of continuous subcutaneous insulin infusion (CSII), does the insulin pump, in addition to improving blood glucose, bring psychological benefits by reducing the pain of multiple daily injections and providing a more flexible lifestyle? Does the frequency of blood glucose monitoring, as an important self-management behavior, reduce anxiety by enhancing the child's sense of "control" over the disease? Furthermore, besides HbA1c, what are the intrinsic links between metabolic indicators such as fasting C-peptide (reflecting residual β-cell function) and acute metabolic events (such as DKA, frequency of hypoglycemia) and anxiety? And are there synergistic effects among treatment methods, management behaviors, and metabolic indicators? These questions have not yet been systematically explored and answered.

[0005] Based on the aforementioned research gaps, there is a lack of existing technologies that can comprehensively consider multiple factors such as metabolic indicators, treatment methods, and management behaviors to make an early and accurate prediction of the risk of anxiety in children with type 1 diabetes mellitus (T1DM).

[0006] Therefore, there is an urgent need in this field for a technical solution that can integrate multifactor analysis, construct predictive models, and provide clinical intervention basis in order to achieve early identification and effective intervention of anxiety in children with T1DM, and improve their physical and mental health and long-term prognosis.

[0007] The information disclosed in this background section is intended only to enhance the understanding of the overall background of the invention and should not be construed as an admission or in any way implying that the information constitutes prior art known to those skilled in the art. Summary of the Invention

[0008] The purpose of this invention is to provide a technical solution that can integrate multi-factor analysis, construct predictive models, and provide clinical intervention basis, so as to achieve early identification and effective intervention of anxiety in children with T1DM, and improve the physical and mental health and long-term prognosis of children.

[0009] To achieve the above objectives, the present invention provides the following solution: A method for predicting the prevalence of anxiety in children with diabetes, comprising the following steps: Collect clinical data of children with diabetes, including demographic data, treatment data, management behavior data, and metabolic index data; The anxiety status of the patients was assessed using the Childhood Anxiety Disorder Screening Scale, with a total SCARED score of ≥23 points as the criterion for determining anxiety status. Based on the clinical data and anxiety state assessment results, univariate analysis, correlation analysis and multivariate logistic regression analysis were performed to construct an anxiety prediction model. The prediction model was used to predict the risk of anxiety in children with diabetes, wherein the multivariate logistic regression model included age, insulin treatment method and glycated hemoglobin as core variables.

[0010] Optionally, the demographic data includes at least one of age, sex, disease duration, weight, height, and body mass index; the treatment method data includes insulin injection method, wherein the insulin injection method is insulin pump or pen injection; the management behavior data includes at least one of blood glucose monitoring frequency, follow-up visit interval, frequency of hypoglycemia, and frequency of diabetic ketoacidosis; the metabolic indicator data includes at least one of glycated hemoglobin, fasting blood glucose, 2-hour postprandial blood glucose, and fasting C-peptide.

[0011] Optionally, the univariate analysis uses t-test, Mann-Whitney U test or chi-square test to screen variables that have statistical differences from anxiety state; the correlation analysis uses Spearman correlation analysis to assess the strength and direction of the association between each variable and the SCARED total score.

[0012] Optionally, in the multivariate logistic regression analysis, age and insulin pump therapy were protective factors against anxiety, while glycated hemoglobin was a risk factor for anxiety; the odds ratio for age was 0.60, the odds ratio for insulin pump therapy was 0.06, and the odds ratio for HbA1c was 2.63.

[0013] Optionally, this further includes using receiver operating characteristic (ROC) curves to assess the predictive power of each clinical indicator for anxiety status, wherein the area under the curve for glycated hemoglobin is ≥0.8, the AUC for DKA frequency is ≥0.7, and the AUC for follow-up visit interval is ≥0.7.

[0014] A predictive system for the prevalence of anxiety in children with diabetes includes: The data acquisition module is used to collect clinical data of children with diabetes, including demographic data, treatment data, management behavior data, and metabolic index data. The psychological assessment module is used to assess the anxiety state of patients using a screening questionnaire for childhood anxiety disorders and output the total SCARED score. The analysis and modeling module is used to perform univariate analysis, correlation analysis and multivariate logistic regression analysis on the clinical data and SCARED total score to build an anxiety prediction model. The risk prediction module is used to output the patient's anxiety risk value based on the prediction model and identify high-risk individuals.

[0015] Optionally, the data acquisition module further includes: The physical examination unit is used to measure height and weight and calculate BMI; The biochemical detection unit is used to detect glycated hemoglobin, fasting blood glucose, 2-hour postprandial blood glucose (2h-PG), and fasting C-peptide. The management and recording unit is used to record insulin treatment methods, blood glucose monitoring frequency, follow-up visit intervals, frequency of hypoglycemia, and frequency of DKA.

[0016] Optionally, when performing multivariate logistic regression analysis, the analysis and modeling module uses age, insulin treatment method, and glycated hemoglobin as core independent variables, and outputs the odds ratio and confidence interval of each variable.

[0017] Optionally, the risk prediction module further includes an ROC analysis unit for calculating the area under the curve of each clinical indicator and outputting a prediction efficacy evaluation based on a preset threshold, wherein the critical value for HbA1c is 9.5%.

[0018] Optionally, the system also includes a report generation module for generating individualized anxiety risk reports and recommending interventions, including insulin pump therapy, optimization of blood glucose monitoring frequency, and adjustment of follow-up visit intervals.

[0019] Compared with the prior art, the present invention has the following beneficial effects: Multi-dimensional data integration: Integrating demographic data, treatment methods, management behaviors, and metabolic indicators to comprehensively assess the anxiety risk in children with diabetes.

[0020] Accurate prediction model: Through univariate analysis, correlation analysis and multivariate logistic regression, a prediction model is constructed that includes age, insulin treatment mode and glycated hemoglobin (HbA1c), which has high predictive efficacy.

[0021] Highly clinically applicable: The model outputs specific risk probabilities, assisting doctors in identifying high-risk children early and providing personalized intervention suggestions (such as insulin pump therapy, optimization of blood glucose monitoring frequency, etc.).

[0022] The system is highly automated: it integrates data collection, psychological assessment, modeling analysis and report generation to achieve intelligent management of the entire process.

[0023] Excellent predictive performance: The ROC curve shows that the AUC of HbA1c is 0.896, and the AUC of DKA frequency and follow-up visit interval are both over 0.7, indicating good model discrimination.

[0024] The intervention is clearly directed: the system not only identifies risks, but also provides specific clinical intervention pathways, which help improve the mental health and blood glucose control of children. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of the method for predicting the prevalence of anxiety in diabetic children according to an embodiment of the present invention.

[0027] Figure 2 This is a schematic diagram of the structure of a system for predicting the prevalence of anxiety in diabetic children, provided in an embodiment of the present invention.

[0028] Figure 3 Receiver operating characteristic (ROC) curves of various clinical indicators for predicting anxiety status in children with type 1 diabetes mellitus (T1DM) provided in this embodiment of the invention. Detailed Implementation

[0029] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] The purpose of this invention is to provide a method and system that can comprehensively consider multiple factors such as metabolic indicators, treatment methods, and management behaviors to make early and accurate predictions of the risk of anxiety in children with type 1 diabetes mellitus (T1DM).

[0031] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0032] Example 1: Specific implementation of a method for predicting the prevalence of anxiety in children with diabetes.

[0033] 1. Data collection and preprocessing: This study included 155 children aged 6-18 years with type 1 diabetes mellitus (T1DM) hospitalized in the Department of Endocrinology of a children's hospital between May 1, 2020 and August 31, 2025. Anxiety status was assessed using the SCARED scale. General information, clinical indicators, and biochemical data were collected, and statistical analysis was performed using t-tests, Mann-Whitney tests, chi-square tests, correlation analysis, logistic regression, and ROC curves. All children met the diagnostic criteria for T1DM and were excluded if they had other chronic diseases, mental illnesses, or recent acute complications. All children were diagnosed according to internationally accepted diagnostic and classification procedures for childhood diabetes to ensure homogeneity and diagnostic accuracy.

[0034] Inclusion criteria for study participants: Children who meet the above diagnostic criteria for TIDM, are aged 6-17 years, have a positive result for diabetes autoantibodies, have normal cognition, are alert and have normal cognitive function, can communicate effectively with researchers, and cooperate in completing all questionnaires; and whose legal guardians understand the research content and voluntarily sign informed consent forms.

[0035] Exclusion criteria for study subjects: (1) Comorbid chronic diseases (such as thyroid dysfunction, Cushing's syndrome, liver dysfunction, cystic fibrosis, celiac disease, epilepsy, autoimmune thyroid disease, etc.); (2) Suffering from congenital diseases, genetic syndromes, or acute or chronic infections; (3) Comorbid intellectual disability, autism spectrum disorder or other mental illnesses that were clearly diagnosed before the diagnosis of diabetes (such as anxiety, depression, bipolar disorder). (4) Currently taking antidepressants, anti-anxiety medications, or other psychotropic drugs; (5) Acute complications such as diabetic ketoacidosis (DKA) or hyperosmolar hyperglycemia have occurred recently (within 1 month prior to enrollment); (6) Unable to cooperate in completing the research questionnaire and data collection; (7) The guardian or the child himself refused to participate in this study.

[0036] The mean age of the study participants was 11.2 ± 2.8 years, and the duration of illness ranged from 1 month to 10 years, with a median duration of 3 (1, 7) years. There were 78 males (50.3%) and 77 females (49.7%). The prevalence of anxiety was 40.3% (63 / 155). Univariate analysis showed that the anxiety group and the non-anxiety group differed significantly in age (10.84±3.07 years vs. 9.78±3.29 years, P=0.044), injection method (higher proportion of pen injections, P=0.016), abnormal injection site (higher proportion, P=0.005), blood glucose monitoring frequency (5.0 [3.0, 8.0] times / day vs. 8.0 [5.0, 11.0] times / day, P=0.001), follow-up visit interval (12.0 [3.0, 24.0] months vs. 6.0 [3.0, 9.0] months, P<0.001), FPG (8.0 [6.5, 9.0] mmol / L vs. 7.0 [6.0, 8.5] mmol / L, P=0.007), and 2h-PG (10.8 [8.5, 12.5] mmol / L vs. 10.5 ... The differences in HbA1c (10.84±2.56% vs. 8.76±2.33%, P<0.001) and the frequency of hypoglycemia (P<0.001) were statistically significant. Multivariate logistic regression showed that age (OR=0.60) and insulin pump use (OR=0.06) were protective factors against anxiety, while high HbA1c (OR=2.63) was a risk factor. ROC analysis indicated that HbA1c (AUC=0.896), DKA frequency (AUC=0.793), and follow-up visit interval (AUC=0.755) had good predictive value for anxiety status.

[0037] Demographic and clinical data of the children were collected using a self-designed questionnaire. The following clinical data were collected: Demographic data: age, sex, height, weight, BMI (Body Mass Index = weight (kg) / height (m)) 2 ()), duration of diabetes, family history, annual family income, annual income of parents and education level of parents.

[0038] Treatment data: Insulin injection method (insulin pump / pen injection), daily insulin dose (u / kg / d).

[0039] Management behavior data: frequency of blood glucose monitoring (times / day), frequency of hypoglycemia (times / day), frequency of DKA (times / year), interval between follow-up visits (months), and whether there are any abnormalities at the injection site.

[0040] Metabolic indicators: fasting plasma glucose (FPG), 2-hour postprandial plasma glucose (2h-PG), glycated hemoglobin (HbA1c), and fasting C-peptide.

[0041] Biochemical tests: All children underwent venous blood collection on the morning of the second day after admission, after fasting for 8 hours. Fasting plasma glucose (FPG), 2-hour postprandial blood glucose (2h-PG), glycated hemoglobin (HbA1c), fasting C-peptide, alanine aminotransferase (ALT), creatinine (Cr), uric acid (UA), total cholesterol (TC), triglycerides (TG), high-density lipoprotein cholesterol (HDL-c), low-density lipoprotein cholesterol (LDL-c), thyroid function (FT3, FT4, TSH), β-hydroxybutyrate, and 25-hydroxyvitamin D were measured using a fully automated biochemical analyzer.

[0042] 2. Anxiety assessment: Anxiety was assessed in children using the Chinese version of the Screening Scale for Childhood Anxiety Disorders (SCARED). The scale consists of 41 items, including five factors: somatization / panic, generalized anxiety, separation anxiety, social phobia, and school phobia. A three-point rating scale (0 (none) - 2 (frequent)) was used, with a total score ≥23 indicating a high probability of anxiety symptoms. In this study, there were 58 children in the anxiety group and 97 children in the non-anxiety group, resulting in an anxiety prevalence rate of 40.3%.

[0043] A database was created using Excel spreadsheets, and statistical analysis was performed using SPSS 26.0 software. Normally distributed continuous data were expressed as mean ± standard deviation (x ± s), and comparisons between groups were performed using the two independent samples t-test. Non-normally distributed continuous data were expressed as median (interquartile range) [M(Q1, Q3)], and comparisons between groups were performed using the Mann-Whitney U test. Categorical data were expressed as number of cases (percentage) [n (%)], and comparisons between groups were performed using the chi-square test. 2 Tests were performed. Pearson or Spearman correlation analysis was used to explore the correlation between anxiety scores and various clinical indicators. Multivariate logistic regression analysis was used to explore independent influencing factors of anxiety, and the odds ratio (OR) and its 95% confidence interval (CI) were calculated. Receiver operating characteristic (ROC) curve analysis was used to analyze the predictive value of each indicator for anxiety status, and the area under the curve (AUC) was calculated. All tests were two-tailed, and P < 0.05 was considered statistically significant.

[0044] 3. Statistical analysis modeling: 3.1 Univariate analysis: The differences in clinical indicators between the anxiety group and the non-anxiety group were compared using the t-test, Mann-Whitney U test, and chi-square test. The results showed that: The anxiety group was older (10.84±3.07 years vs. 9.78±3.29 years, P=0.044).

[0045] The anxiety group had a higher rate of insulin pen use (56.9% vs. 38.1%, P=0.016).

[0046] The anxiety group had higher HbA1c levels (10.84±2.56% vs. 8.76±2.33%, P<0.001).

[0047] The anxiety group had a lower frequency of blood glucose monitoring (5.0 [3.0, 8.0] times / day vs. 8.0 [5.0, 11.0] times / day, P=0.001).

[0048] 3.2 Correlation Analysis: Spearman correlation analysis was used, and the results showed that: The total SCARED score was positively correlated with HbA1c (r=0.391, P=0.002).

[0049] The total SCARED score was negatively correlated with insulin pump use (r=-0.261, P=0.048).

[0050] The total SCARED score was negatively correlated with the frequency of blood glucose monitoring (r=-0.297, P=0.023).

[0051] 3.3 Multivariate Logistic Regression Analysis: Variables with p < 0.1 in the univariate analysis were included in the multivariate logistic regression model. The final variables included in the model are: Age (OR=0.60, 95%CI: 0.411~0.888, P=0.010) was a protective factor.

[0052] Insulin pump therapy (OR=0.06, 95%CI: 0.024~0.153, P<0.001) was a strong protective factor.

[0053] HbA1c (OR=2.63, 95%CI: 1.760~3.918, P<0.001) was a risk factor.

[0054] 4. Predictive model construction and validation: Based on the results of multivariate logistic regression, an anxiety prediction model was constructed: Logit(P) = β0 + β1 × age + β2 × insulin pump + β3 × HbA1c.

[0055] in: Insulin pump therapy is coded as 1, and pen injection is coded as 0.

[0056] The model coefficients were set based on the regression results: β1=-0.504, β2=-2.796, β3=0.966.

[0057] 5. ROC curve analysis and predictive performance evaluation: Plot ROC curves to assess the predictive value of each indicator for anxiety states: The AUC of HbA1c was 0.896 (95% CI: 0.845–0.946), with an optimal cutoff value of 9.5%.

[0058] The AUC for DKA frequency was 0.793 (95% CI: 0.718–0.868).

[0059] The AUC for the interval between follow-up visits was 0.755 (95% CI: 0.675–0.835).

[0060] Comparison of general clinical data and metabolic indicators between children with anxiety and those without anxiety: This study included 155 children with type 1 diabetes mellitus (T1DM), aged 6–18 years, with a mean age of 11.2 ± 2.8 years; disease duration ranged from 1 month to 10 years, with a median duration of 3 (1, 7) years. Among them, 78 were male (50.3%) and 77 were female (49.7%). Based on their Scared scores, they were divided into an anxiety group (SCARED total score ≥ 23, n=58) and a non-anxiety group (SCARED total score < 23, n=97). The prevalence of anxiety was 40.3% (63 / 155). The comparison results of general information and clinical metabolic indicators between the two groups are detailed in Table 1.

[0061] Regarding demographic characteristics, the mean age of children in the anxiety group was (10.84±3.07) years, significantly higher than that in the non-anxiety group (9.78±3.29) years (t=2.029, P=0.044). However, there were no statistically significant differences between the two groups in terms of sex ratio (male / female: 31 / 27 vs. 52 / 45), weight, height, and body mass index (BMI) (all P>0.05).

[0062] Disease management indicators showed significant differences in treatment methods between the two groups (x 2 =5.83, P=0.016. The proportion of children using insulin pens was higher in the anxiety group (33 / 58, 56.9%), while the proportion using insulin pumps was higher in the non-anxiety group (60 / 97, 61.9%). The proportion of injection site abnormalities (such as redness, swelling, and induration) was significantly higher in the anxiety group than in the non-anxiety group (39 / 58 vs. 44 / 97, x=5.83, P=0.016). 2=7.78, P=0.005). The median frequency of daily blood glucose monitoring in the anxiety group was significantly lower than that in the non-anxiety group [5.0 (3.0, 8.0) times / day vs. 8.0 (5.0, 11.0) times / day, U=1760, P=0.001], and the interval between follow-up visits was longer [12.0 (3.0, 24.0) months vs. 6.0 (3.0, 9.0) months, U=1865, P<0.001]. There were no statistically significant differences between the two groups in terms of diabetes duration, family history of diabetes, and average annual number of DKA events.

[0063] Comparison of metabolic control indicators revealed key differences. Children in the anxiety group had significantly worse glycemic control than those in the non-anxiety group. Their fasting plasma glucose (FPG) [8.0 (6.5, 9.0) mmol / L vs. 7.0 (6.0, 8.5) mmol / L, U=2102.5, P=0.007], 2-hour postprandial glucose (2h-PG) [10.8 (8.5, 12.5) mmol / L vs. 8.9 (7.8, 11.6) mmol / L, U=1946.5, P=0.001] and glycated hemoglobin (HbA1c) levels [(10.84 ± 2.56)% vs. (8.76 ± 2.33)%, t=5.318, P<0.001] were all significantly higher. The number of daily hypoglycemic episodes was significantly higher in the anxiety group than in the non-anxiety group (P<0.001).

[0064] Regarding pancreatic function and therapeutic dosage, there were no statistically significant differences between the two groups in fasting C-peptide levels, fasting insulin levels, and daily insulin dosage per unit body weight (all P>0.05).

[0065] In summary, children with type 1 diabetes mellitus who exhibit anxiety tend to be older, more likely to use pen injection therapy, have a higher incidence of injection site abnormalities, lower frequency of blood glucose monitoring, longer intervals between follow-up visits, and poorer blood glucose control (manifested as higher FPG, 2h-PG, HbA1c, and more frequent hypoglycemic events).

[0066] Table 1. Comparison of general information and clinical metabolic indicators between the anxiety group and the non-anxiety group. To further explore the strength and direction of the association between various clinical variables and anxiety, this study used Spearman correlation analysis to analyze the total SCARED score and various indicators. The results are detailed in Table 2.

[0067] The analysis results showed that multiple disease management and metabolic control indicators were statistically significantly associated with the anxiety levels of the children.

[0068] Significant positive correlation (P<0.05): The study found a positive correlation between the anxiety score (SCARED total score) of the children and the duration of diabetes (r=0.382, P=0.003), indicating that the longer the duration of the disease, the heavier the burden of disease management and the more pronounced the anxiety. Anxiety level was also positively correlated with injection site abnormalities (such as induration and lipomatosis) (r=0.324, P=0.013), suggesting that local complications caused by treatment are a significant source of psychological stress. Most importantly, anxiety was significantly associated with metabolic control indicators, including glycated hemoglobin (HbA1c) (r=0.391, P=0.002) and 2-hour postprandial blood glucose (2h-PG) (r=0.368, P=0.005), strongly suggesting a close link between poor glycemic control and anxiety. Furthermore, anxiety was positively correlated with the interval between follow-up visits (r=0.421, P=0.001), meaning that the longer the interval between visits, the higher the anxiety level in children. This may reflect that the lack of regular medical support increases patients' sense of insecurity. More noteworthy is the strong correlation between the frequency of acute complications and anxiety. The frequency of DKA (digestive ketoacidosis) (r=0.446, P<0.001) and hypoglycemia (r=0.463, P<0.001) were both strongly positively correlated with anxiety scores, indicating that repeated occurrences of acute metabolic events are a strong factor causing or exacerbating anxiety.

[0069] Significant negative correlation (P<0.05): This study also identified some potential protective factors. Injection method (insulin pump = 1, pen injection = 0) was negatively correlated with anxiety score (r = -0.261, P = 0.048), indicating that the use of the more advanced insulin pump treatment technology is associated with lower anxiety levels. Frequent blood glucose monitoring was negatively correlated with anxiety score (r = -0.297, P = 0.023), suggesting that more proactive self-monitoring may reduce anxiety by enhancing the sense of control over the condition. Furthermore, fasting C-peptide levels were also negatively correlated with anxiety score (r = -0.278, P = 0.034), indicating that preserving some endogenous insulin secretion function may be associated with better psychological adaptability.

[0070] No statistically significant correlations were found: The analysis also showed that age, sex, weight, height, BMI, fasting blood glucose (FBG), fasting insulin level, and total daily insulin usage were not significantly correlated with the children's anxiety levels (P>0.05). A family history of diabetes showed a positive trend with anxiety scores, but this correlation did not reach statistical significance (r=0.236, P=0.073).

[0071] In summary, the correlation analysis in this study showed that anxiety in children with type 1 diabetes mellitus (T1DM) was positively correlated with disease duration, treatment discomfort (abnormal injection site), glycemic control level (high HbA1c, high 2h-PG), frequency of acute metabolic events (DKA, hypoglycemia), and insufficient medical support (long intervals between follow-up visits); while it was negatively correlated with better treatment techniques (insulin pump), proactive management behaviors (frequent blood glucose monitoring), and better residual pancreatic function (high fasting C-peptide).

[0072] Table 2. Correlation analysis between anxiety scores and clinical indicators in children with type 1 diabetes mellitus (T1DM)

[0073] Multivariate logistic regression analysis of factors influencing anxiety in children with type 1 diabetes mellitus (T1DM): To control for interference among variables and identify independent influencing factors of anxiety, we included variables that were statistically significant or nearly significant in the univariate analysis into a multivariate binary logistic regression model (forward: LR method). Anxiety (SCARED total score ≥ 23 points) was used as the dependent variable. The analysis results are detailed in Table 3.

[0074] The three variables that were ultimately included in the regression model were age, insulin treatment method, and glycated hemoglobin (HbA1c).

[0075] Age: The analysis showed that age was an independent protective factor against anxiety (OR=0.60, 95% CI: 0.411–0.888, P=0.010). This result indicates that, after adjusting for treatment methods and blood glucose levels, the risk of anxiety decreased by 40% for each year of age increase (1 / 0.60≈1.67, i.e., the risk was approximately 0.6 times the original risk). This seems to contradict the common clinical observation that children around puberty experience increased psychological stress, suggesting that after excluding the interference of factors such as blood glucose control and technical treatment, the cognitive maturity and disease adaptation associated with increasing age may play a dominant protective role.

[0076] Insulin therapy method: Insulin pump use was the strongest independent protective factor against anxiety (OR=0.06, 95%CI: 0.024~0.153, P<0.001). Compared with insulin pen injection, children using insulin pump therapy had a 94% lower risk of developing anxiety (1-0.06=0.94). This result remained highly significant after excluding the effects of glycemic control level (HbA1c) and other confounding factors, strongly suggesting that insulin pump therapy itself—potentially by reducing the number of daily injections, providing more flexible lifestyles, improving quality of life, and enhancing patients' sense of control over their condition—has a unique and significant positive effect on alleviating anxiety.

[0077] Glycated hemoglobin (HbA1c): HbA1c was an independent risk factor for anxiety (OR=2.63, 95% CI: 1.760–3.918, P<0.001). This result indicates that, after adjusting for age and treatment, for every 1% increase in HbA1c, the risk of anxiety in children increased 2.63-fold. This confirms a close and independent intrinsic link between poor long-term glycemic control and anxiety, potentially forming a vicious cycle.

[0078] Variables not included in the final model: Notably, several variables associated with anxiety in univariate or correlation analyses, such as disease duration, frequency of hypoglycemia, injection site abnormalities, intervals between follow-up visits, frequency of blood glucose monitoring, and fasting C-peptide, did not show independent predictive value in multivariate analysis (all P>0.05). This suggests that these variables may be indirectly associated with anxiety primarily by influencing the three core factors mentioned above (especially HbA1c) or by collinearity with other variables, rather than being independent direct influencing factors.

[0079] Conclusion: Multivariate logistic regression analysis ultimately identified advanced age and insulin pump therapy as independent protective factors against anxiety in children with type 1 diabetes mellitus (T1DM), while high HbA1c levels were an independent risk factor. This finding focuses interventions on promoting advanced treatment technologies (insulin pumps) and optimizing glycemic control (lowering HbA1c), providing clear and precise targets for clinical strategies to prevent and alleviate anxiety.

[0080] Table 3. Multivariate Logistic Regression Analysis of Factors Influencing Anxiety in Children with Type 1 Diabetes Mellitus Note: The dependent variable is anxiety level (SCARED total score ≥ 23). The injection method variable is defined as "pen". Statistical significance is set at α = 0.05, with bold indicating statistical significance. OR: odds ratio; CI: confidence interval.

[0081] Assessment of the predictive value of various clinical indicators for anxiety status in children with type 1 diabetes mellitus (T1DM) To assess the ability of different clinical indicators to identify anxiety states, this study plotted receiver operating characteristic (ROC) curves and calculated the area under the curve (AUC). The results are detailed in Table 4. Figure 3 .

[0082] A superior predictive indicator: Glycated hemoglobin (HbA1c) demonstrated exceptional predictive power for anxiety, with an AUC of 0.896 (95% CI: 0.845–0.946, P<0.001). This result indicates that HbA1c, as a continuous and objective metabolic indicator, can very accurately distinguish between anxious and non-anxious children, making it the strongest biomarker for predicting anxiety risk. Based on the Youden index, its optimal cutoff value is approximately 9.5%; above this value, the risk of anxiety in children increases significantly.

[0083] Good predictive indicators: Annual frequency of DKA (AUC=0.793, P<0.001) and length of follow-up visit interval (AUC=0.755, P<0.001) also showed good predictive power. Both indicators reflect the serious instability of disease management and the lack of medical support. Frequent DKA attacks mean repeated occurrences of acute metabolic crises, while long follow-up visit intervals may indicate insufficient disease monitoring and professional guidance; both are important clinical situations that trigger or exacerbate anxiety in children.

[0084] Clinically significant negative predictors (protective factors): This study identified two variables with significantly lower AUC values ​​than 0.5: insulin treatment method (pump vs. pen) (AUC = 0.233, P < 0.001) and frequency of blood glucose monitoring (AUC = 0.283, P < 0.001). This does not statistically indicate a lack of predictive ability, but rather signifies a strong negative correlation between these variables and anxiety levels, meaning they are powerful protective factors. This result further strongly confirms that insulin pump use and more frequent blood glucose monitoring are associated with a very low risk of anxiety in children.

[0085] 4. Indicators with limited or no predictive value: While 2-hour postprandial blood glucose (2h-PG) was statistically significant (AUC=0.614, P=0.020), its discriminative ability was actually poor, limiting its clinical value when used alone. Furthermore, demographic indicators such as age and height, as well as other metabolic indicators such as FBG and C-peptide, showed no statistically significant difference in AUC values ​​compared to 0.5 (P>0.05), indicating that these indicators, as single markers, cannot effectively differentiate the anxiety state of children.

[0086] Conclusion: ROC curve analysis reinforced the core findings of this study from a predictive perspective: HbA1c is the most critical indicator for screening anxiety risk, and routine psychological assessments should be performed on children with HbA1c > 9.5% in clinical practice. A history of DKA and long intervals between follow-up visits are important early warning signals of anxiety risk. Insulin pump therapy and high-frequency blood glucose monitoring are not only treatment methods but also highly valuable psychological interventions, and their protective effects have been quantitatively confirmed. Future research could consider combining HbA1c, treatment methods, and management behavior indicators to construct a comprehensive psychological intervention system. A comprehensive predictive model is developed to enable early and accurate identification of anxiety risk in children.

[0087] Table 4. Predictive efficacy analysis of various clinical indicators on anxiety status in children with type 1 diabetes mellitus (T1DM) Note: AUC: Area Under the Curve. Null Hypothesis: True AUC = 0.5. The p-value represents the result of the test of the null hypothesis. Bold indicates p < 0.05, indicating that the discriminative power of the variable is considered statistically significant. The analysis is performed under nonparametric assumptions. AUC: Area Under the Curve, used to assess the discriminative power of predictors. Judgment Criteria: 0.5 (None), 0.7-0.8 (Moderate), 0.8-0.9 (Good), >0.9 (Excellent). The p-value is derived from the test of the null hypothesis "True AUC = 0.5". P < 0.05 indicates that the discriminative power of the variable is statistically significant. Bold indicates p < 0.05, which is statistically significant. The analysis is performed under nonparametric assumptions. Some variables have tied values ​​between groups, and the statistical data may have slight bias.

[0088] This study investigated the synergistic effects of metabolic indicators, treatment methods, and management behaviors on anxiety in children with type 1 diabetes mellitus (T1DM) using a cross-sectional survey system. Results showed that 40.3% of children with T1DM exhibited anxiety symptoms, significantly higher than in the general child population. Multivariate analysis indicated that high HbA1c levels were an independent risk factor for anxiety (OR=2.63), while insulin pump use (OR=0.06) and increasing age (OR=0.60) were protective factors. ROC analysis further confirmed that HbA1c had excellent predictive power for anxiety (AUC=0.896). These findings provide important theoretical and practical evidence for the comprehensive management of children with T1DM.

[0089] The two-way relationship between blood sugar control and anxiety: This study confirms that HbA1c is the strongest predictor of anxiety, consistent with the findings of Butwicka et al. The mechanism may involve multiple levels: physiologically, chronic hyperglycemia can activate the hypothalamic-pituitary-adrenal axis (HPA axis), promoting the secretion of stress hormones such as cortisol, directly affecting the central nervous system's mood regulation function. Simultaneously, the inflammatory response and oxidative stress induced by hyperglycemia may also cross the blood-brain barrier, affecting the function of the prefrontal cortex and limbic system, brain regions closely related to mood regulation. Behavioralally, poor glycemic control is often accompanied by frequent blood glucose fluctuations and hypoglycemic events. Children and their families experience "diabetes distress" due to concerns about acute complications, and this chronic psychological stress can easily translate into anxiety. The positive correlation between hypoglycemic frequency and anxiety found in this study (r=0.463) supports this mechanism.

[0090] Psychological benefits brought about by innovative treatment techniques: This study found that insulin pump therapy is the strongest protective factor against anxiety, a finding with significant clinical implications. Compared to traditional multiple daily injections, insulin pumps provide more stable glycemic control and reduce glycemic fluctuations, thereby alleviating children's fear of hypoglycemia. More importantly, pump therapy significantly improves treatment adherence and lifestyle flexibility, reducing the disease's disruption to daily life. This "de-stigmatization" effect is particularly important for improving the psychological adaptation of adolescent patients. This finding provides strong psychological support for the promotion of advanced treatment technologies.

[0091] The key role of managerial behavior: This study revealed a negative correlation between the frequency of blood glucose monitoring and anxiety levels (r=-0.297). Frequent monitoring may reduce anxiety by enhancing children's sense of "control" over their condition, a finding consistent with self-efficacy theory. Simultaneously, the study found a positive correlation between follow-up visit intervals and anxiety (r=0.421), suggesting that longer intervals may indicate a lack of medical support, increasing patients' insecurity and loneliness. These findings highlight the importance of continuous medical support and self-management education in psychological interventions.

[0092] New findings on the age-protective effect: Contrary to common belief, this study found that age was a protective factor after controlling blood sugar and treatment methods. This may be because children with the disease gradually mature their disease coping abilities and self-management skills as they grow older

[10] . At the same time, older children may receive better psychological support through peer support and social networks. This finding suggests that we should pay attention to the psychological adaptation problems of younger children with the disease and develop age-specific psychological intervention programs.

[0093] The innovativeness and limitations of the study: The innovations of this study lie in: the first-ever construction of a comprehensive predictive model incorporating metabolic indicators, treatment methods, and management behaviors; the use of multivariate analysis to reveal the independent effects among various factors; and the quantification of the predictive value of each indicator using ROC curves, providing a practical tool for clinical screening. However, this study has several limitations: the cross-sectional design cannot establish causal relationships; the single-center nature of the study may affect the generalizability of the results; and potential influencing factors such as family environment and social support were not included. Future research should employ a longitudinal design, incorporating more sociopsychological variables to further validate the findings of this study.

[0094] This study confirms that anxiety in children with type 1 diabetes mellitus (T1DM) is closely related to metabolic control, treatment methods, and management behaviors. HbA1c is a key indicator for anxiety screening, while insulin pump therapy and frequent blood glucose monitoring are important protective factors. In clinical practice, a comprehensive "bio-psychological-behavioral" intervention model should be established, incorporating psychological assessment into routine follow-up, promoting advanced treatment technologies, and strengthening self-management education to achieve dual management of physical and mental health.

[0095] 6. Clinical application examples: A 12-year-old child with type 1 diabetes mellitus (T1DM) was treated with an insulin pen and had an HbA1c level of 10.5%. Substitute this data into the predictive model: Logit(P)=β0-0.504×12-2.796×0+0.966×10.5; The calculated P value is greater than 0.5, indicating that the individual is classified as a high-risk anxiety individual.

[0096] Intervention recommendations: Switching to insulin pump therapy is recommended.

[0097] Strengthen blood glucose monitoring, aiming for ≥8 times / day.

[0098] Shorten the interval between follow-up visits to ≤6 months.

[0099] Regular SCARED assessments should be conducted, and referral to a psychiatric department should be made if necessary.

[0100] Example 2: Specific implementation of a predictive system for the prevalence of anxiety in children with diabetes: 1. System Architecture: This system includes the following modules: The system includes a data acquisition module, a psychological assessment module, an analysis and modeling module, a risk prediction module, and a report generation module.

[0101] 2. Module Function Details: The data acquisition module includes: Physical examination unit: measures height and weight, and automatically calculates BMI.

[0102] Biochemical detection unit: Connects to the LIS system to automatically collect indicators such as HbA1c, FPG, 2h-PG, and C-peptide.

[0103] Management record unit: Enter insulin treatment method, blood glucose monitoring frequency, follow-up visit interval, etc.

[0104] Psychological assessment module: The system integrates an electronic version of the SCARED scale, allowing children to complete the assessment on a tablet.

[0105] The system automatically calculates the total score, and marks a score of ≥23 as an anxiety state.

[0106] Analysis and modeling module: Built-in statistical analysis engine, automatically performs: Univariate analysis (t-test / U-test / ) x ²Test) Spearman correlation analysis; Multivariate Logistic Regression Analysis; Output the OR value and confidence interval for the core variables (age, insulin pump, HbA1c).

[0107] Risk prediction module: ROC analysis unit: Calculates the AUC of each indicator, with a focus on monitoring HbA1c (critical value 9.5%).

[0108] Risk calculation unit: Outputs the probability of individual anxiety risk based on the prediction model.

[0109] High-risk warning: When the risk probability is greater than 50%, the system will automatically mark it and remind medical staff.

[0110] The report generation module is used to generate personalized anxiety risk reports, including: Risk level (low / medium / high) Analysis of key influencing factors; Personalized intervention recommendations (such as recommending insulin pumps, optimizing monitoring frequency, etc.).

[0111] System application example: Case: A 14-year-old female patient with a 5-year history of illness, using an insulin pen, with an HbA1c level of 11.2%, and blood glucose monitoring 4 times a day, with follow-up visits every 12 months.

[0112] System processing flow: The data acquisition module inputs clinical data.

[0113] The psychological assessment module completes the SCARED assessment (total score 26 points).

[0114] The analysis and modeling module identified the core variables: age (14 years), treatment method (pen), and HbA1c (11.2%).

[0115] The risk prediction module calculates the probability of anxiety risk as 72%, and marks it as high risk.

[0116] The report generation module outputs a report; suggestions: Immediately switch to insulin pump therapy; Increase blood glucose monitoring to ≥7 times / day; The interval between follow-up visits has been shortened to 3 months; Arrange a consultation with the Department of Psychology.

[0117] 4. System advantages: Integrate multi-source data to achieve automated collection and analysis.

[0118] The model is accurate and reliable, and was built based on large-sample clinical studies.

[0119] It has strong clinical applicability, provides specific intervention measures, and helps individualized comprehensive management.

[0120] This embodiment constructs an anxiety prediction model incorporating age, insulin treatment method, and HbA1c based on clinical data from 155 children with type 1 diabetes mellitus (T1DM), and develops a corresponding prediction system. ROC curve validation shows that the model has excellent predictive efficacy (AUC = 0.896 for HbA1c), possessing high clinical application value and providing an effective tool for early identification and intervention of anxiety in children with T1DM.

[0121] This system reveals the pathogenesis and multiple influencing factors of anxiety in children with type 1 diabetes mellitus (T1DM). It innovatively constructs an anxiety prediction model based on metabolic indicators and treatment methods, highlighting the psychological protective effects of insulin pump therapy and frequent blood glucose monitoring, and providing empirical evidence and clinical prediction tools for psychological intervention in children with T1DM.

[0122] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the descriptions are relatively simple; relevant parts can be referred to the method section.

[0123] This document uses specific examples to illustrate the principles and implementation methods of the present invention. The descriptions of the above embodiments are only for the purpose of helping to understand the method and core ideas of the present invention. Furthermore, those skilled in the art will recognize that, based on the ideas of the present invention, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A method for predicting the prevalence of anxiety in children with diabetes, characterized in that, Includes the following steps: Collect clinical data of children with diabetes, including demographic data, treatment data, management behavior data, and metabolic index data; The anxiety status of the patients was assessed using the Childhood Anxiety Disorder Screening Scale, with a total SCARED score of ≥23 points as the criterion for determining anxiety status. Based on the clinical data and anxiety state assessment results, univariate analysis, correlation analysis and multivariate logistic regression analysis were performed to construct an anxiety prediction model. The prediction model was used to predict the risk of anxiety in children with diabetes, wherein the multivariate logistic regression model included age, insulin treatment method and glycated hemoglobin as core variables.

2. The method for predicting the prevalence of anxiety in diabetic children according to claim 1, characterized in that, The demographic data includes at least one of age, sex, disease duration, weight, height, and body mass index; the treatment data includes insulin injection method, wherein the insulin injection method is insulin pump or pen injection; the management behavior data includes at least one of blood glucose monitoring frequency, follow-up visit interval, frequency of hypoglycemia, and frequency of diabetic ketoacidosis; the metabolic indicator data includes at least one of glycated hemoglobin, fasting blood glucose, 2-hour postprandial blood glucose, and fasting C-peptide.

3. The method for predicting the prevalence of anxiety in diabetic children according to claim 1, characterized in that, The univariate analysis used t-tests, Mann-Whitney U tests, or chi-square tests to screen variables that were statistically different from anxiety states; the correlation analysis used Spearman correlation analysis to assess the strength and direction of the association between each variable and the SCARED total score.

4. The method for predicting the prevalence of anxiety in diabetic children according to claim 1, characterized in that, In the multivariate logistic regression analysis, age and insulin pump therapy were protective factors against anxiety, while glycated hemoglobin (HbA1c) was a risk factor for anxiety; the odds ratio for age was 0.60, the odds ratio for insulin pump therapy was 0.06, and the odds ratio for HbA1c was 2.

63.

5. The method for predicting the prevalence of anxiety in diabetic children according to claim 1, characterized in that, Further, the study included using receiver operating characteristic (ROC) curves to assess the predictive power of each clinical indicator for anxiety status, with the area under the curve for glycated hemoglobin ≥0.8, the AUC for DKA frequency ≥0.7, and the AUC for follow-up visit interval ≥0.

7.

6. A predictive system for the prevalence of anxiety in children with diabetes, characterized in that, include: The data acquisition module is used to collect clinical data of children with diabetes, including demographic data, treatment data, management behavior data, and metabolic index data. The psychological assessment module is used to assess the anxiety state of patients using a screening questionnaire for childhood anxiety disorders and output the total SCARED score. The analysis and modeling module is used to perform univariate analysis, correlation analysis and multivariate logistic regression analysis on the clinical data and SCARED total score to build an anxiety prediction model. The risk prediction module is used to output the patient's anxiety risk value based on the prediction model and identify high-risk individuals.

7. The predictive system for the prevalence of anxiety in diabetic children according to claim 6, characterized in that, The data acquisition module further includes: The physical examination unit is used to measure height and weight and calculate BMI; The biochemical detection unit is used to detect glycated hemoglobin, fasting blood glucose, 2-hour postprandial blood glucose (2h-PG), and fasting C-peptide. The management and recording unit is used to record insulin treatment methods, blood glucose monitoring frequency, follow-up visit intervals, frequency of hypoglycemia, and frequency of DKA.

8. The predictive system for the prevalence of anxiety in diabetic children according to claim 6, characterized in that, When performing multivariate logistic regression analysis, the analysis and modeling module uses age, insulin treatment method, and glycated hemoglobin as core independent variables and outputs the odds ratio and confidence interval of each variable.

9. The predictive system for the prevalence of anxiety in diabetic children according to claim 6, characterized in that, The risk prediction module further includes an ROC analysis unit, which is used to calculate the area under the curve of each clinical indicator and output a prediction efficacy evaluation based on a preset threshold, wherein the critical value of HbA1c is 9.5%.

10. The predictive system for the prevalence of anxiety in diabetic children according to claim 6, characterized in that, The system also includes a report generation module for generating individualized anxiety risk reports and recommending interventions, including insulin pump therapy, optimization of blood glucose monitoring frequency, and adjustment of follow-up visit intervals.

Images