Machine Learning-Based Prediction of Hypoglycemia Severity in Hospitalized Diabetic Patients

Machine Learning-Based Prediction of Hypoglycemia Severity in Hospitalized Diabetic Patients | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Machine Learning-Based Prediction of Hypoglycemia Severity in Hospitalized Diabetic Patients Hongjian Jia¹, Jietao Zhang¹ This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6619286/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Objective: To identify risk factors for hypoglycemia in hospitalized patients with type 2 diabetes mellitus (T2DM) and develop predictive models for hypoglycemia severity based on machine learning algorithms. Methods: Adult non-pregnant hospitalized patients diagnosed with T2DM were retrospectively enrolled from the electronic medical record system of the Affiliated Hospital of Qingdao University. Patients were categorized into hypoglycemia groups (mild, moderate-to-severe) or a non-hypoglycemia group based on inpatient venous plasma glucose levels. After data preprocessing, univariate and multivariate analyses were conducted to identify significant predictors. Three predictive models (XGBoost, Random Forest [RF], and Logistic Regression) were subsequently constructed and validated to evaluate their predictive performances. Results: From an initial cohort of 8,947 patients, 1,798 patients were included after data screening. Among the evaluated models, the RF model demonstrated the highest predictive accuracy (93.3%) and Kappa coefficient (0.873), followed by XGBoost (accuracy: 92.6%, Kappa: 0.860). Logistic regression exhibited comparatively lower performance (accuracy: 83.8%, Kappa: 0.685). The macro-average area under the ROC curve (AUC) values for RF, XGBoost, and logistic regression were 0.960, 0.955, and 0.788, respectively, highlighting the superior discriminative capability of the RF model. While both XGBoost and RF models identified glycemic control metrics and glucose variability as core predictors for hypoglycemia, the RF model additionally emphasized medication usage, whereas XGBoost prioritized basal metabolic parameters. Conclusions: The RF model outperformed XGBoost and conventional logistic regression in predicting hypoglycemia severity among hospitalized T2DM patients. The results emphasize the importance of closely monitoring glucose levels and glucose variability during diabetes management to prevent hypoglycemia. The developed model provides a foundation for implementing preventive strategies to reduce hypoglycemia occurrence in hospitalized patients with T2DM. Health sciences/Diseases Health sciences/Endocrinology Health sciences/Medical research Type 2 diabetes mellitus Hypoglycemia Risk prediction Machine learning Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction Diabetes mellitus (DM) is a major chronic disease worldwide, with prevalence rates steadily rising over recent decades [ 1 ] . Diabetes and its associated complications pose significant threats to patient health and quality of life, greatly impacting patients' daily activities and potentially leading to mortality in severe cases. Among these complications, hypoglycemia is particularly critical to prevent and manage due to its frequent occurrence and substantial health risks for diabetic patients [ 2 ] . Hypoglycemic events in diabetic patients are multifactorial, influenced by medications, diet, lifestyle, and comorbidities [ 3 – 5 ] . Recently, stringent glycemic control strategies have been associated with an increased risk of hypoglycemia. However, hypoglycemia can undermine the long-term benefits gained from good glycemic management, underscoring the importance of carefully balancing the benefits and risks of intensive glucose management. Hospitals play a central role in glycemic control, adjustments of antidiabetic medications, and individualized care for diabetic patients, placing heightened responsibilities on healthcare professionals. Patients with diabetes commonly exhibit multiple comorbidities that contribute independently to hypoglycemia risk. Studies have identified older age, renal impairment, liver dysfunction, poor nutritional status, inappropriate medication use, and debilitating diseases as critical risk factors for hypoglycemia. Clinical practitioners thus must analyze these risk factors systematically to accurately assess the probability of hypoglycemia and implement effective preventive measures. Predicting hypoglycemia accurately, however, remains challenging due to the complex interplay of various clinical and biological factors. Although multiple studies have confirmed insulin therapy [ 6 ] , impaired renal function, and suboptimal glycemic control as significant predictors of hypoglycemia, reliably forecasting such events remains difficult. Traditional logistic regression models have been widely employed to identify risk factors but are limited by their assumption of linear relationships between predictors and outcomes. With advancements in machine learning techniques, models such as Extreme Gradient Boosting (XGBoost) and Random Forest (RF) have shown remarkable performance in predicting complex clinical outcomes due to their ability to capture nonlinear interactions and handle high-dimensional data. Nevertheless, few studies have systematically compared traditional statistical methods and machine learning algorithms in predicting hypoglycemia severity among hospitalized patients with diabetes. Thus, the current study aimed to develop and compare three predictive models—multinomial logistic regression, XGBoost, and RF—to identify independent risk factors and predict the severity of hypoglycemia among hospitalized patients with type 2 diabetes mellitus (T2DM). Additionally, we performed a feature importance analysis to provide clinical insights into the most critical predictors associated with hypoglycemia risk. Our findings may facilitate early intervention strategies and enhance individualized patient management during hospitalization. Machine Learning Model Development Three predictive models were developed and evaluated: multinomial logistic regression, XGBoost, and Random Forest. Hyperparameters for the XGBoost and RF models were optimized using cross-validation. Multiclass ROC curves were constructed using a One-vs-Rest approach to visualize and compare the models' predictive capabilities. Methods Study Design and Population This retrospective study was conducted using data from the electronic medical record system of the Affiliated Hospital of Qingdao University. Initially, 8,947 adult, non-pregnant hospitalized patients with a confirmed diagnosis of T2DM were identified. After applying inclusion and exclusion criteria, a total of 1,798 patients were included in the final analysis. Data Collection Clinical and demographic data were extracted from electronic medical records, including demographic characteristics, clinical features, laboratory findings, and antidiabetic medication use. Variables collected were age, gender, body mass index (BMI) classification, Charlson Comorbidity Index (CCI), glycated hemoglobin (HbA1c), mean blood glucose levels, serum creatinine, C-peptide, lipid profile, and use of antidiabetic medications (e.g., insulin, metformin, DPP-4 inhibitors). Patient Grouping Patients were categorized into three groups based on venous plasma glucose levels measured during hospitalization: Normal glycemia: >3.9 mmol/L Mild hypoglycemia: 3.0–3.9 mmol/L Moderate-to-severe hypoglycemia: <3.0 mmol/L Statistical Analysis Univariate analyses were conducted using the chi-square test for categorical variables and the Kruskal-Wallis test for continuous variables. Variables with a P-value < 0.05 were entered into a multivariate multinomial logistic regression analysis. Model performance was evaluated using overall accuracy, Kappa statistic, the area under the receiver operating characteristic (ROC) curve (AUC), and confusion matrices. All statistical analyses were performed using R software and SPSS. Results Patient Characteristics A total of 1,798 hospitalized diabetic patients were included in the final analysis. Based on inpatient blood glucose measurements, patients were divided into three groups: normoglycemic, mild hypoglycemia, and moderate-to-severe hypoglycemia. Significant differences were observed among the three groups in age, Charlson Comorbidity Index (CCI), the number of glucose-lowering medication classes, triglycerides (TG), serum creatinine, glycated hemoglobin (HbA1c), C-peptide, mean glucose levels, cholesterol, gender distribution, BMI classification, and use of various glucose-lowering medications (including SGLT2 inhibitors, α-glucosidase inhibitors, metformin, thiazolidinediones, and insulin) (all P < 0.05). Baseline characteristics of the study population are detailed in Table 1 . Table 1 Baseline characteristics of patients by hypoglycemia severity group Variables Total (n = 1798) Normal (n = 1046) Mild hypoglycemia (n = 593) Moderate-to-severe hypoglycemia (n = 159) P Age, M (Q₁, Q₃) 62.00 (54.00, 68.00) 61.00 (54.00,67.00) 63.00 (54.00,69.00) 62.00 (56.00,71.00) 0.012 Charlson, M (Q₁, Q₃) 1.00 (1.00, 1.00) 1.00 (1.00,1.00) 1.00 (1.00,3.00) 2.00 (1.00,3.00) < .001 DrugClassCount, M (Q₁, Q₃) 2.00 (1.00, 3.00) 2.00 (1.00,3.00) 3.00 (2.00,3.00) 2.00 (1.00,3.00) < .001 TG, M (Q₁, Q₃) 1.10 (0.77, 1.67) 1.15 (0.79,1.75) 1.04 (0.75,1.48) 1.02 (0.68,1.63) < .001 Creatinine, M (Q₁, Q₃) 76.60 (57.00, 93.00) 73.00 (55.00,90.25) 79.50 (59.00,97.00) 90.00 (64.80,133.22) < .001 HbA1c, M (Q₁, Q₃) 6.20 (5.90, 7.30) 6.10 (5.80,6.30) 7.70 (6.50,9.50) 7.60 (6.50,9.45) < .001 C Peptide, M (Q₁, Q₃) 1.67 (0.70, 2.67) 2.10 (1.45,3.04) 0.69 (0.33,1.63) 0.73 (0.24,2.00) < .001 GlucoseAvg, M (Q₁, Q₃) 4.92 (3.84, 5.63) 5.19 (4.78,5.65) 3.73 (3.48,4.45) 4.77 (2.79,8.49) < .001 Cholesterol, M (Q₁, Q₃) 4.25 (3.42, 5.03) 4.21 (3.39,4.95) 4.35 (3.51,5.28) 4.06 (3.28,4.98) 0.005 Sex, n(%) 0.014 M 999 (55.56) 609 (58.22) 301 (50.76) 89 (55.97) W 799 (44.44) 437 (41.78) 292 (49.24) 70 (44.03) BMI Class, n(%) < .001 Overweight 708 (39.38) 439 (41.97) 227 (38.28) 42 (26.42) Obese 339 (18.85) 232 (22.18) 88 (14.84) 19 (11.95) Underweight 49 (2.73) 14 (1.34) 22 (3.71) 13 (8.18) Normal 702 (39.04) 361 (34.51) 256 (43.17) 85 (53.46) DPP4, n(%) 0.869 Not used 1063 (59.12) 615 (58.80) 351 (59.19) 97 (61.01) Used 735 (40.88) 431 (41.20) 242 (40.81) 62 (38.99) GLP1, n(%) 0.056 Not used 1753 (97.50) 1012 (96.75) 584 (98.48) 157 (98.74) Used 45 (2.50) 34 (3.25) 9 (1.52) 2 (1.26) SGLT2, n(%) 0.027 Not used 1560 (86.76) 894 (85.47) 518 (87.35) 148 (93.08) Used 238 (13.24) 152 (14.53) 75 (12.65) 11 (6.92) AGI, n(%) < .001 Not used 796 (44.27) 519 (49.62) 214 (36.09) 63 (39.62) Used 1002 (55.73) 527 (50.38) 379 (63.91) 96 (60.38) Metformin, n(%) < .001 Not used 740 (41.16) 346 (33.08) 298 (50.25) 96 (60.38) Used 1058 (58.84) 700 (66.92) 295 (49.75) 63 (39.62) TZD, n(%) 0.013 Not used 1700 (94.55) 975 (93.21) 571 (96.29) 154 (96.86) Used 98 (5.45) 71 (6.79) 22 (3.71) 5 (3.14) Insulin, n(%) < .001 Not used 931 (51.78) 811 (77.53) 96 (16.19) 24 (15.09) Used 867 (48.22) 235 (22.47) 497 (83.81) 135 (84.91) Univariate and Multivariate Logistic Regression Analyses In the univariate logistic regression analysis (Table 2 ), gender, Charlson comorbidity index, number of glucose-lowering medication classes, BMI classification, triglycerides (TG), serum creatinine, HbA1c, C-peptide, mean glucose levels, cholesterol, and use of specific hypoglycemic drugs (metformin, α-glucosidase inhibitors, insulin) were significantly associated with the risk of hypoglycemia (P < 0.05). Specifically, an increased Charlson index, lower BMI, reduced C-peptide levels, decreased TG, elevated HbA1c, elevated creatinine levels, and insulin use were linked to a higher risk of hypoglycemia. Table 2 Univariate multinomial logistic regression analysis for predictors of hypoglycemia severity Variables Mild hypoglycemia vs Normal (OR [95%CI]) P value Moderate-to-severe hypoglycemia vs Normal (OR [95%CI]) P value Sex (Male vs Female) 0.74 (0.60–0.91) 0.004 0.91 (0.65–1.28) 0.593 Age (per year) 1.01 (1.00–1.02) 0.156 1.01 (1.00–1.03) 0.133 Charlson score 3.81 (3.18–4.56) < 0.001 4.07 (3.35–4.94) < 0.001 Drug Class Count 1.59 (1.44–1.75) < 0.001 1.28 (1.09–1.51) 0.002 BMI Overweight vs Normal 0.73 (0.58–0.91) 0.006 0.41 (0.27–0.60) < 0.001 BMI Obese vs Normal 0.54 (0.40–0.72) < 0.001 0.35 (0.21–0.59) < 0.001 BMI Underweight vs Normal 2.22 (1.11–4.41) 0.024 3.94 (1.79–8.70) 0.001 TG (mmol/L) 0.85 (0.77–0.94) 0.002 0.95 (0.83–1.09) 0.454 HbA1c (%) 10.15 (7.76–13.27) < 0.001 10.07 (7.65–13.26) < 0.001 Creatinine (umol/L) 1.005 (1.003–1.007) < 0.001 1.006 (1.004–1.008) < 0.001 GlucoseAvg (mmol/L) 0.77 (0.72–0.83) < 0.001 1.11 (1.04–1.18) 0.003 C-Peptide (ng/mL) 0.68 (0.63–0.74) < 0.001 0.87 (0.80–0.95) 0.002 Cholesterol (mmol/L) 1.15 (1.07–1.23) < 0.001 0.99 (0.88–1.11) 0.811 DPP4i (No vs Yes) 1.02 (0.83–1.25) 0.876 1.10 (0.78–1.54) 0.597 GLP1-RA (No vs Yes) 2.18 (1.04–4.58) 0.039 2.64 (0.63–11.09) 0.186 SGLT2i (No vs Yes) 1.17 (0.87–1.58) 0.289 2.29 (1.21–4.32) 0.011 AGI (No vs Yes) 0.57 (0.47–0.70) < 0.001 0.67 (0.47–0.94) 0.019 Metformin (No vs Yes) 2.04 (1.66–2.51) < 0.001 3.08 (2.19–4.34) < 0.001 TZD (No vs Yes) 1.89 (1.16–3.08) 0.011 2.24 (0.89–5.64) 0.086 Insulin (No vs Yes) 0.06 (0.04–0.07) < 0.001 0.05 (0.03–0.08) < 0.001 In the multivariate multinomial logistic regression analysis (Table 3 ), independent factors associated with mild hypoglycemia included older age (OR = 0.966, 95% CI: 0.951–0.981, P < 0.001), higher Charlson comorbidity index (OR = 2.684, 95% CI: 2.121–3.398, P < 0.001), increased creatinine (OR = 1.005, 95% CI: 1.003–1.007, P < 0.001), elevated HbA1c (OR = 10.570, 95% CI: 7.554–14.789, P < 0.001), lower mean glucose levels (OR = 0.545, 95% CI: 0.479–0.619, P < 0.001), and insulin use (OR = 0.205, 95% CI: 0.127–0.332, P < 0.001). For moderate-to-severe hypoglycemia, independent predictors included older age (OR = 0.973, 95% CI: 0.954–0.993, P = 0.007), elevated Charlson comorbidity index (OR = 2.744, 95% CI: 2.141–3.517, P < 0.001), increased serum creatinine (OR = 1.005, 95% CI: 1.003–1.007, P < 0.001), elevated HbA1c (OR = 10.047, 95% CI: 7.127–14.163, P < 0.001), lower mean glucose levels (OR = 0.633, 95% CI: 0.554–0.723, P < 0.001), and insulin use (OR = 0.170, 95% CI: 0.086–0.336, P < 0.001). Additionally, overweight status was a protective factor against moderate-to-severe hypoglycemia (OR = 0.511, 95% CI: 0.302–0.865, P = 0.012), while underweight status significantly increased the risk (OR = 3.654, 95% CI: 1.045–12.781, P = 0.043). Overall, the multivariate logistic regression model demonstrated good fit (χ² = 1585.430, df = 40, P < 0.001; Cox-Snell R² = 0.586; Nagelkerke R² = 0.703; McFadden R² = 0.492). Table 3 Multivariate multinomial logistic regression results for mild and moderate-to-severe hypoglycemia Variables Mild hypoglycemia vs Normal (OR [95%CI]) P value Moderate-to-Severe hypoglycemia vs Normal (OR [95%CI]) P value Age 0.966 (0.951–0.981) < 0.001 0.973 (0.954–0.993) 0.007 Charlson Score 2.684 (2.121–3.398) < 0.001 2.744 (2.141–3.517) < 0.001 Drug Class Count 0.788 (0.571–1.086) 0.146 0.791 (0.520–1.202) 0.272 TG 0.857 (0.712–1.033) 0.105 0.904 (0.717–1.139) 0.391 Creatinine 1.005 (1.003–1.007) < 0.001 1.005 (1.003–1.007) < 0.001 HbA1c 10.570 (7.554–14.789) < 0.001 10.047 (7.127–14.163) < 0.001 C-Peptide 1.023 (0.945–1.108) 0.570 1.092 (0.994–1.200) 0.067 Glucose Avg 0.545 (0.479–0.619) < 0.001 0.633 (0.554–0.723) < 0.001 Cholesterol 1.021 (0.894–1.166) 0.758 0.856 (0.728–1.006) 0.059 Sex (Male) 0.995 (0.692–1.431) 0.980 1.086 (0.682–1.730) 0.728 SGLT2 Not Used 0.953 (0.500–1.817) 0.883 1.809 (0.714–4.580) 0.211 AGI Not Used 0.680 (0.425–1.089) 0.109 0.680 (0.362–1.277) 0.230 Metformin Not Used 1.034 (0.628–1.701) 0.895 1.162 (0.607–2.223) 0.651 TZD Not Used 1.640 (0.673–3.999) 0.277 1.601 (0.475–5.390) 0.447 Insulin Not Used 0.205 (0.127–0.332) < 0.001 0.170 (0.086–0.336) < 0.001 BMI Class - Overweight 1.043 (0.699–1.557) 0.836 0.511 (0.302–0.865) 0.012 BMI Class - Obese 1.254 (0.765–2.057) 0.370 0.747 (0.380–1.470) 0.399 BMI Class - Underweight 1.964 (0.600–6.427) 0.265 3.654 (1.045–12.781) 0.043 Machine Learning Model Evaluation The predictive performances of multinomial logistic regression, XGBoost, and random forest (RF) models were evaluated using receiver operating characteristic (ROC) curves constructed through a One-vs-Rest approach. ROC curves for each model are illustrated individually in Fig. 1 ,Fig. 2 ,Fig. 3 , and a direct comparative analysis among the three models is presented in Fig. 4 . The multinomial logistic regression model yielded area under the curve (AUC) values of 0.938 for predicting normoglycemia, 0.889 for mild hypoglycemia, and 0.754 for moderate-to-severe hypoglycemia. The XGBoost model demonstrated notably higher AUCs of 0.991, 0.980, and 0.943, respectively. Similarly, the RF model achieved AUCs of 0.991, 0.983, and 0.952 for the three respective severity categories. Comparative ROC analysis (Fig. 4 ) clearly indicated superior discriminative performance of both XGBoost and RF models across all hypoglycemia severity categories compared to logistic regression. Model Performance Comparison To comprehensively evaluate predictive performances, we compared the overall accuracy, Kappa coefficient, and macro-average AUC of the three models (Table 4 ). The RF model achieved the highest accuracy (93.3%) and Kappa coefficient (0.873), followed closely by XGBoost (accuracy: 92.6%, Kappa: 0.860). Logistic regression demonstrated relatively lower performance, with an accuracy of 83.8% and Kappa of 0.685. Consistently, the macro-average AUC values for RF, XGBoost, and logistic regression were 0.960, 0.955, and 0.788, respectively, further emphasizing the superior predictive capability of the tree-based machine learning algorithms. Table 4 Performance comparison of three models: Accuracy, AUC, and Kappa Model Accuracy Kappa Macro AUC Logistic Regression 0.838 0.6845 0.7878 XGBoost 0.9255 0.8603 0.9553 Random Forest 0.9330 0.8729 0.9596 Feature Importance Analysis Feature importance analysis (Fig. 5 ) identified mean blood glucose level and HbA1c as the two most critical predictors in both XGBoost and RF models. Subsequent important predictors in the XGBoost model included creatinine, C-peptide, and triglycerides (TG). In the RF model, following mean glucose and HbA1c, important predictors included C-peptide, insulin usage, and the Charlson comorbidity index. Discussion In the present study, we developed and systematically compared three predictive models—multinomial logistic regression, XGBoost, and random forest (RF)—to identify key predictors of hypoglycemia severity among hospitalized patients with diabetes. Our results demonstrated that both machine learning models (XGBoost and RF) exhibited superior discriminative performance compared to traditional logistic regression, especially for detecting moderate-to-severe hypoglycemia, this has some modeling studies with the same results [ 7 ] . We utilized the One-vs-Rest strategy, a standard and effective approach for generating ROC curves in multiclass classification problems, allowing robust evaluation and comparison of each model's predictive performance across hypoglycemia categories. The dataset used to construct our predictive models comprised common clinical and laboratory variables routinely available for hospitalized patients with type 2 diabetes mellitus (T2DM). The logistic regression model identified seven statistically significant predictors (P < 0.05), while both XGBoost and RF models highlighted the top 10 most influential variables. Notably, six predictors—age, Charlson Comorbidity Index (CCI), serum creatinine, HbA1c, mean glucose levels, and insulin use—emerged consistently as significant factors across all three models. A number of previous modeling studies and factor analysis studies have similarly given the same conclusions for one or more of these factors [ 8 – 10 ] . While previous studies have individually confirmed the significance of one or several of these factors, our research uniquely integrated multiple models for comprehensive comparison, offering a broader perspective than single-model studies previously reported. A key advantage of our study lies in its reliance on routinely collected clinical data. This facilitates rapid assessment of hypoglycemia risk without additional specialized testing, thus granting medical staff valuable time to implement preventive strategies and appropriate treatments. Regarding predictive variables, HbA1c has consistently been identified as a crucial factor in hypoglycemia research. Prior studies have reported that both excessively high HbA1c levels (> 9%) and overly stringent glycemic control (HbA1c < 7%) are associated with increased hypoglycemia risk [ 11 ] . Additionally, insulin use within these HbA1c ranges further exacerbates hypoglycemia risk. Both our univariate and multivariate analyses supported these findings, reinforcing that elevated HbA1c levels and insulin therapy are critical risk factors for hypoglycemia in diabetic patients. In recent years, growing attention has focused on the association between the Charlson Comorbidity Index (CCI) and hypoglycemia. Most prior studies investigating comorbidities typically examined only one or a few components of the CCI, such as cardiovascular disease, renal impairment, or malignancy [ 12 ] . These comorbid conditions have been associated with glucose instability and heightened risk of adverse glycemic events. Patients with multiple chronic diseases often experience altered drug metabolism, polypharmacy, and malnutrition, potentially interfering with glucose regulation and insulin sensitivity. Incorporating the CCI into all three predictive models in our study enhances its utility as a comprehensive clinical indicator. Unlike individual diagnoses, the CCI provides an aggregated measure of overall disease burden, capturing complex interactions between comorbidities and thereby improving the generalizability and interpretability of the predictive models. Integrating the CCI into predictive assessments may assist clinicians in accurately stratifying risk, particularly among elderly hospitalized patients or those with multiple chronic diseases, who inherently have higher susceptibility to hypoglycemic episodes [ 13 ] . Early identification of high-CCI patients could facilitate personalized monitoring, medication adjustments, and tailored nutritional plans, thereby mitigating hypoglycemia risk. Lastly, we conducted feature importance analyses for the top ten variables in the XGBoost and RF models. Both models consistently identified glycemic control and glucose variability as key predictors. Additionally, XGBoost emphasized basal metabolic parameters (e.g., creatinine and C-peptide), whereas the RF model placed greater emphasis on medication use. These findings highlight nuanced differences between machine learning approaches, underscoring their ability to provide targeted clinical insights into managing and preventing hypoglycemic events in hospitalized diabetic patients. Conclusions In this study, we developed and compared three prediction models - multinomial logistic regression, XGBoost, and random forest - to identify factors associated with hypoglycemia severity in hospitalized diabetic patients. Our results showed that both machine learning models outperformed the traditional logistic regression model in terms of predictive performance, with higher accuracy and AUC values in all hypoglycemia categories. This time, the three models were designed with all commonly used clinical indicators, without the need for special tests, to prevent hypoglycemia as early as possible in type 2 diabetic patients and to avoid the adverse consequences caused by hypoglycemia. In a comparison of the three models and the focus of the models, the six variables of age, Charlson comorbidity index, creatinine, glycosylated hemoglobin, mean blood glucose level, and insulin use were consistently identified as core predictors in almost all models [ 14 – 16 ] , This further enhances their clinical relevance。Future studies should focus on the prospective validation and practical application of these models to assess their clinical utility and impact on patient safety. Limitations The model was designed to use only common clinical indicators, and because hypoglycaemia occurs in a variety of ways, even if the model had a higher degree of accuracy, it would still not be able to fully diagnose all types of hypoglycaemia. As this study was based on single-centre data, further validation of the predictive effect of the model is needed in the future. Conflict of interest statement The authors declare that they have no financial conflict of interest with regard to the content of this report. Declarations The study titled “Machine Learning-Based Prediction of Hypoglycemia Severity in Hospitalized Diabetic Patients” was reviewed and approved by the Ethics Committee of The Affiliated Hospital of Qingdao University. All procedures involving human participants were conducted in accordance with institutional guidelines and approved protocols. Author Contribution Hongjian Jia contributed to study design, data preprocessing, modeling, and manuscript writing and is the first author.Jietao Zhang supervised the project, interpreted clinical significance, and is the corresponding author. References ONG K L, STAFFORD L K, MCLAUGHLIN, S. A. et al. Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: a systematic analysis for the Global Burden of Disease Study 2021[J]. Lancet 402 (10397), 203–234. 10.1016/S0140-6736(23)01301-6 (2023). NUGRAHA A R A, Y. U. N. I. R. E. et al. Risk factors of severe hypoglycemia among patients with type 2 diabetes mellitus in outpatient clinic of tertiary hospital in Indonesia[J]. Sci. Rep. 13 (1), 16259. 10.1038/s41598-023-43459-2 (2023). BONAVENTURA, A., MONTECUCCO, F. & DALLEGRI, F. Update on strategies limiting iatrogenic hypoglycemia[J/OL]. (2015-09-01)[2025-05-01]. https://ec.bioscientifica.com/view/journals/ec/4/3/R37.xml . (2015). 10.1530/EC-15-0044 Anonymous Predictors of hypoglycemia in hospitalized patients with diabetes mellitus | Internal and Emergency Medicine[EB/OL]. [2025-05-01]. https://link.springer.com/article/10.1007/s11739-018-1787-0 Anonymous. Risk factors for hypoglycemia in patients with type 2 diabetes, hospitalized in internal medicine wards: Findings from the FADOI-DIAMOND study - Diabetes Research and Clinical Practice[EB/OL]. [2025-05-01]. https://www.diabetesresearchclinicalpractice.com/article/S0168-8227(16)00069-3/abstract Anonymous Rates and predictors of hypoglycaemia in 27 585 people from 24 countries with insulin-treated type 1 and type 2 diabetes: the global HAT study - Khunti – 2016 - Diabetes, Obesity and Metabolism - Wiley Online Library[EB/OL]. [2025-05-01]. https://dom-pubs.onlinelibrary.wiley.com/doi/full/10.1111/dom.12689 Anonymous Interdisciplinary Nursing Research[EB/OL]. [2025-05-01]. https://journals.lww.com/inr/fulltext/2025/03000/machine_learning_based_prediction_model_for.4.aspx Anonymous The Risk Factors of Severe Hypoglycemia in Older Patients with Dementia and Type 2 Diabetes Mellitus - PMC[EB/OL]. [2025-05-01]. https://pmc.ncbi.nlm.nih.gov/articles/PMC8779381/?utm_source=chatgpt.com#sec5-jpm-12-00067 Anonymous Hypoglycemia in patients with type 2 diabetes mellitus during hospitalization: associated factors and prognostic value | Diabetology & Metabolic Syndrome | Full Text[EB/OL]. [2025-05-01]. https://dmsjournal.biomedcentral.com/articles/ 10.1186/s13098-023-01212-9?utm_source=chatgpt.com#Sec8 AKIROV, A. et al. Predictors of hypoglycemia in hospitalized patients with diabetes mellitus[J]. Intern. Emerg. Med. 13 (3), 343–350. 10.1007/s11739-018-1787-0 (2018). LIPSKA K J et al. HbA1c and Risk of Severe Hypoglycemia in Type 2 Diabetes[J]. Diabetes Care . 36 (11), 3535–3542. 10.2337/dc13-0610 (2013). Anonymous. The risk factors of inpatient hypoglycemia: A systematic review - PMC[EB/OL]. [2025-05-01]. https://pmc.ncbi.nlm.nih.gov/articles/PMC7218453/?utm_source=chatgpt.com#sec3 Anonymous Prevalence and predictors of hypoglycemia in older outpatients with type 2 diabetes mellitus | PLOS One[EB/OL]. [2025-05-01]. https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0309618&utm_source=chatgpt.com#sec016 Anonymous. Predicting hypoglycemia in elderly inpatients with type 2 diabetes: the ADOCHBIU model - PMC[EB/OL]. [2025-05-01]. https://pmc.ncbi.nlm.nih.gov/articles/PMC11602273/?utm_source=chatgpt.com Anonymous. Enhancing severe hypoglycemia prediction in type 2 diabetes mellitus through multi-view co-training machine learning model for imbalanced dataset | Scientific Reports[EB/OL]. [2025-05-01]. https://www.nature.com/articles/s41598-024-69844-z?utm_source=chatgpt.com Anonymous. Predicting the Risk of Inpatient Hypoglycemia With Machine Learning Using Electronic Health Records. | Diabetes Care | American Diabetes Association[EB/OL]. [2025-05-01]. https://diabetesjournals.org/care/article/43/7/1504/35530/Predicting-the-Risk-of-Inpatient-Hypoglycemia-With?utm_source=chatgpt.com Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6619286","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":453713660,"identity":"524555a4-7361-4b90-b1e2-ec242ed2d3d6","order_by":0,"name":"Hongjian Jia¹","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAzUlEQVRIiWNgGAWjYBACxhkMiQ8SDCR4+NkbGx9+IFJLssGHChsZyZ7DzcYSRFkjwcAmOeNMmo3BjfQ2AR5idDDPbnggzdt2mEdy5sM2oH47Od0GQg6bcyDBGKSFXzqx7UEBQ7Kx2QFCWmYkJCSDbZmd2G4gwXAgcRsxWg6DtBjcPNgmwUOklsRGoPd5DG4wEq8lmQEYyDySPYnAQDYgwi+GM3LSfwCj0p6f/fjDhx8q7OQIa2ngSUDiGhBQDgLyDOyETB0Fo2AUjIIRDwAnNUacfgBAFAAAAABJRU5ErkJggg==","orcid":"","institution":"The Affiliated Hospital of Qingdao University","correspondingAuthor":true,"prefix":"","firstName":"Hongjian","middleName":"","lastName":"Jia¹","suffix":""},{"id":453713661,"identity":"b4d50b74-e812-40d7-9174-22191af8833c","order_by":1,"name":"Jietao Zhang¹","email":"","orcid":"","institution":"The Affiliated Hospital of Qingdao University","correspondingAuthor":false,"prefix":"","firstName":"Jietao","middleName":"","lastName":"Zhang¹","suffix":""}],"badges":[],"createdAt":"2025-05-08 09:53:22","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6619286/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6619286/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":82892272,"identity":"204d44b5-bfb3-4f8f-82f7-179f6498c4c4","added_by":"auto","created_at":"2025-05-16 12:18:59","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":51324,"visible":true,"origin":"","legend":"\u003cp\u003eOne-vs-rest ROC curves for multinomial logistic regression model\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6619286/v1/e6bcf161687e6876914bbb7d.png"},{"id":82892268,"identity":"4e6a2a34-a093-41a0-90ed-d9b5abbc42a6","added_by":"auto","created_at":"2025-05-16 12:18:59","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":46746,"visible":true,"origin":"","legend":"\u003cp\u003eOne-vs-rest ROC curves for the XGBoost model\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6619286/v1/19601077f05f6eee2630d095.png"},{"id":82892269,"identity":"9bdc9c9b-e95c-4676-a9d3-5d726460069e","added_by":"auto","created_at":"2025-05-16 12:18:59","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":50437,"visible":true,"origin":"","legend":"\u003cp\u003eOne-vs-rest ROC curves for random forest model\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6619286/v1/cec6e31e10c498690c34dd69.png"},{"id":82893667,"identity":"5896e276-0d4b-48bc-9fc3-3a4bae3ce9e2","added_by":"auto","created_at":"2025-05-16 12:26:59","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":97929,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of ROC curves across logistic regression, XGBoost, and random forest models\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6619286/v1/f11994d3cc7952bf0d6b8f70.png"},{"id":82895378,"identity":"51634e70-ea0d-4eed-8ac9-ccb163043ec1","added_by":"auto","created_at":"2025-05-16 12:42:59","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":21796,"visible":true,"origin":"","legend":"\u003cp\u003eTop 10 Feature Importance\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6619286/v1/1ef01dae220c7bbbf2eccda3.png"},{"id":83341536,"identity":"c11732d3-00c3-4556-8790-761154cf2693","added_by":"auto","created_at":"2025-05-23 10:46:45","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1060240,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6619286/v1/64e8a9ae-6f22-4fe8-a01c-7e6af2e3fd83.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Machine Learning-Based Prediction of Hypoglycemia Severity in Hospitalized Diabetic Patients","fulltext":[{"header":"Introduction","content":"\u003cp\u003eDiabetes mellitus (DM) is a major chronic disease worldwide, with prevalence rates steadily rising over recent decades\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]\u003c/sup\u003e. Diabetes and its associated complications pose significant threats to patient health and quality of life, greatly impacting patients' daily activities and potentially leading to mortality in severe cases. Among these complications, hypoglycemia is particularly critical to prevent and manage due to its frequent occurrence and substantial health risks for diabetic patients \u003csup\u003e[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]\u003c/sup\u003e. Hypoglycemic events in diabetic patients are multifactorial, influenced by medications, diet, lifestyle, and comorbidities \u003csup\u003e[\u003cspan additionalcitationids=\"CR4\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e–\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]\u003c/sup\u003e. Recently, stringent glycemic control strategies have been associated with an increased risk of hypoglycemia. However, hypoglycemia can undermine the long-term benefits gained from good glycemic management, underscoring the importance of carefully balancing the benefits and risks of intensive glucose management.\u003c/p\u003e \u003cp\u003eHospitals play a central role in glycemic control, adjustments of antidiabetic medications, and individualized care for diabetic patients, placing heightened responsibilities on healthcare professionals. Patients with diabetes commonly exhibit multiple comorbidities that contribute independently to hypoglycemia risk. Studies have identified older age, renal impairment, liver dysfunction, poor nutritional status, inappropriate medication use, and debilitating diseases as critical risk factors for hypoglycemia. Clinical practitioners thus must analyze these risk factors systematically to accurately assess the probability of hypoglycemia and implement effective preventive measures.\u003c/p\u003e \u003cp\u003ePredicting hypoglycemia accurately, however, remains challenging due to the complex interplay of various clinical and biological factors. Although multiple studies have confirmed insulin therapy\u003csup\u003e[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e]\u003c/sup\u003e, impaired renal function, and suboptimal glycemic control as significant predictors of hypoglycemia, reliably forecasting such events remains difficult. Traditional logistic regression models have been widely employed to identify risk factors but are limited by their assumption of linear relationships between predictors and outcomes.\u003c/p\u003e \u003cp\u003eWith advancements in machine learning techniques, models such as Extreme Gradient Boosting (XGBoost) and Random Forest (RF) have shown remarkable performance in predicting complex clinical outcomes due to their ability to capture nonlinear interactions and handle high-dimensional data. Nevertheless, few studies have systematically compared traditional statistical methods and machine learning algorithms in predicting hypoglycemia severity among hospitalized patients with diabetes.\u003c/p\u003e \u003cp\u003eThus, the current study aimed to develop and compare three predictive models—multinomial logistic regression, XGBoost, and RF—to identify independent risk factors and predict the severity of hypoglycemia among hospitalized patients with type 2 diabetes mellitus (T2DM). Additionally, we performed a feature importance analysis to provide clinical insights into the most critical predictors associated with hypoglycemia risk. Our findings may facilitate early intervention strategies and enhance individualized patient management during hospitalization.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec2\" class=\"Section2\"\u003e \u003cp\u003eMachine Learning Model Development\u003c/p\u003e \u003cp\u003eThree predictive models were developed and evaluated: multinomial logistic regression, XGBoost, and Random Forest. Hyperparameters for the XGBoost and RF models were optimized using cross-validation. Multiclass ROC curves were constructed using a One-vs-Rest approach to visualize and compare the models' predictive capabilities.\u003c/p\u003e \u003c/div\u003e"},{"header":"Methods","content":"\u003cp\u003eStudy Design and Population\u003c/p\u003e\u003cp\u003eThis retrospective study was conducted using data from the electronic medical record system of the Affiliated Hospital of Qingdao University. Initially, 8,947 adult, non-pregnant hospitalized patients with a confirmed diagnosis of T2DM were identified. After applying inclusion and exclusion criteria, a total of 1,798 patients were included in the final analysis.\u003c/p\u003e\u003cp\u003eData Collection\u003c/p\u003e\u003cp\u003eClinical and demographic data were extracted from electronic medical records, including demographic characteristics, clinical features, laboratory findings, and antidiabetic medication use. Variables collected were age, gender, body mass index (BMI) classification, Charlson Comorbidity Index (CCI), glycated hemoglobin (HbA1c), mean blood glucose levels, serum creatinine, C-peptide, lipid profile, and use of antidiabetic medications (e.g., insulin, metformin, DPP-4 inhibitors).\u003c/p\u003e\u003cp\u003ePatient Grouping\u003c/p\u003e\u003cp\u003ePatients were categorized into three groups based on venous plasma glucose levels measured during hospitalization:\u003c/p\u003e\u003cp\u003e \u003c/p\u003e\u003cul\u003e \u003cli\u003e \u003cp\u003eNormal glycemia: \u0026gt;3.9 mmol/L\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eMild hypoglycemia: 3.0–3.9 mmol/L\u003c/p\u003e \u003c/li\u003e \u003cli\u003e \u003cp\u003eModerate-to-severe hypoglycemia: \u0026lt;3.0 mmol/L\u003c/p\u003e \u003c/li\u003e \u003c/ul\u003e\u003cp\u003e\u003c/p\u003e\u003ch2\u003eStatistical Analysis\u003c/h2\u003e\u003cp\u003eUnivariate analyses were conducted using the chi-square test for categorical variables and the Kruskal-Wallis test for continuous variables. Variables with a P-value \u0026lt; 0.05 were entered into a multivariate multinomial logistic regression analysis. Model performance was evaluated using overall accuracy, Kappa statistic, the area under the receiver operating characteristic (ROC) curve (AUC), and confusion matrices. All statistical analyses were performed using R software and SPSS.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003ePatient Characteristics\u003c/p\u003e \u003cp\u003eA total of 1,798 hospitalized diabetic patients were included in the final analysis. Based on inpatient blood glucose measurements, patients were divided into three groups: normoglycemic, mild hypoglycemia, and moderate-to-severe hypoglycemia. Significant differences were observed among the three groups in age, Charlson Comorbidity Index (CCI), the number of glucose-lowering medication classes, triglycerides (TG), serum creatinine, glycated hemoglobin (HbA1c), C-peptide, mean glucose levels, cholesterol, gender distribution, BMI classification, and use of various glucose-lowering medications (including SGLT2 inhibitors, α-glucosidase inhibitors, metformin, thiazolidinediones, and insulin) (all P\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Baseline characteristics of the study population are detailed in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eBaseline characteristics of patients by hypoglycemia severity group\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"6\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTotal (n\u0026thinsp;=\u0026thinsp;1798)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNormal (n\u0026thinsp;=\u0026thinsp;1046)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMild hypoglycemia (n\u0026thinsp;=\u0026thinsp;593)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eModerate-to-severe hypoglycemia (n\u0026thinsp;=\u0026thinsp;159)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eP\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge, M (Q₁, Q₃)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e62.00 (54.00, 68.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e61.00 (54.00,67.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e63.00 (54.00,69.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e62.00 (56.00,71.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.012\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharlson, M (Q₁, Q₃)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.00 (1.00, 1.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.00 (1.00,1.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.00 (1.00,3.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.00 (1.00,3.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDrugClassCount, M (Q₁, Q₃)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.00 (1.00, 3.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2.00 (1.00,3.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.00 (2.00,3.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2.00 (1.00,3.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTG, M (Q₁, Q₃)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.10 (0.77, 1.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1.15 (0.79,1.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.04 (0.75,1.48)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e1.02 (0.68,1.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCreatinine, M (Q₁, Q₃)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e76.60 (57.00, 93.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e73.00 (55.00,90.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e79.50 (59.00,97.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e90.00 (64.80,133.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHbA1c, M (Q₁, Q₃)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e6.20 (5.90, 7.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e6.10 (5.80,6.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e7.70 (6.50,9.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e7.60 (6.50,9.45)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eC Peptide, M (Q₁, Q₃)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.67 (0.70, 2.67)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e2.10 (1.45,3.04)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.69 (0.33,1.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.73 (0.24,2.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGlucoseAvg, M (Q₁, Q₃)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4.92 (3.84, 5.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e5.19 (4.78,5.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.73 (3.48,4.45)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4.77 (2.79,8.49)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCholesterol, M (Q₁, Q₃)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e4.25 (3.42, 5.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e4.21 (3.39,4.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4.35 (3.51,5.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e4.06 (3.28,4.98)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.014\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eM\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e999 (55.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e609 (58.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e301 (50.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e89 (55.97)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eW\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e799 (44.44)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e437 (41.78)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e292 (49.24)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e70 (44.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI Class, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOverweight\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e708 (39.38)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e439 (41.97)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e227 (38.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e42 (26.42)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eObese\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e339 (18.85)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e232 (22.18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e88 (14.84)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e19 (11.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUnderweight\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e49 (2.73)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e14 (1.34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e22 (3.71)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e13 (8.18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNormal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e702 (39.04)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e361 (34.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e256 (43.17)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e85 (53.46)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDPP4, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.869\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1063 (59.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e615 (58.80)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e351 (59.19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e97 (61.01)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUsed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e735 (40.88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e431 (41.20)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e242 (40.81)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e62 (38.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGLP1, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.056\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1753 (97.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e1012 (96.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e584 (98.48)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e157 (98.74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUsed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e45 (2.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e34 (3.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e9 (1.52)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e2 (1.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSGLT2, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.027\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1560 (86.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e894 (85.47)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e518 (87.35)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e148 (93.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUsed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e238 (13.24)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e152 (14.53)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e75 (12.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e11 (6.92)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAGI, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e796 (44.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e519 (49.62)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e214 (36.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e63 (39.62)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUsed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1002 (55.73)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e527 (50.38)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e379 (63.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e96 (60.38)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetformin, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e740 (41.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e346 (33.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e298 (50.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e96 (60.38)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUsed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1058 (58.84)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e700 (66.92)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e295 (49.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e63 (39.62)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZD, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e0.013\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1700 (94.55)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e975 (93.21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e571 (96.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e154 (96.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUsed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e98 (5.45)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e71 (6.79)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e22 (3.71)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e5 (3.14)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInsulin, n(%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNot used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e931 (51.78)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e811 (77.53)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e96 (16.19)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e24 (15.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eUsed\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e867 (48.22)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e235 (22.47)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e497 (83.81)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e135 (84.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eUnivariate and Multivariate Logistic Regression Analyses\u003c/p\u003e \u003cp\u003eIn the univariate logistic regression analysis (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e), gender, Charlson comorbidity index, number of glucose-lowering medication classes, BMI classification, triglycerides (TG), serum creatinine, HbA1c, C-peptide, mean glucose levels, cholesterol, and use of specific hypoglycemic drugs (metformin, α-glucosidase inhibitors, insulin) were significantly associated with the risk of hypoglycemia (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05). Specifically, an increased Charlson index, lower BMI, reduced C-peptide levels, decreased TG, elevated HbA1c, elevated creatinine levels, and insulin use were linked to a higher risk of hypoglycemia.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eUnivariate multinomial logistic regression analysis for predictors of hypoglycemia severity\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMild hypoglycemia vs Normal (OR [95%CI])\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eP value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eModerate-to-severe hypoglycemia vs Normal (OR [95%CI])\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eP value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex (Male vs Female)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.74 (0.60\u0026ndash;0.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.91 (0.65\u0026ndash;1.28)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.593\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge (per year)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.01 (1.00\u0026ndash;1.02)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.156\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.01 (1.00\u0026ndash;1.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.133\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharlson score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e3.81 (3.18\u0026ndash;4.56)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e4.07 (3.35\u0026ndash;4.94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDrug Class Count\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.59 (1.44\u0026ndash;1.75)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.28 (1.09\u0026ndash;1.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI Overweight vs Normal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.73 (0.58\u0026ndash;0.91)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.006\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.41 (0.27\u0026ndash;0.60)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI Obese vs Normal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.54 (0.40\u0026ndash;0.72)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.35 (0.21\u0026ndash;0.59)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI Underweight vs Normal\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.22 (1.11\u0026ndash;4.41)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.024\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.94 (1.79\u0026ndash;8.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTG (mmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.85 (0.77\u0026ndash;0.94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.95 (0.83\u0026ndash;1.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.454\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHbA1c (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10.15 (7.76\u0026ndash;13.27)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10.07 (7.65\u0026ndash;13.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCreatinine (umol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.005 (1.003\u0026ndash;1.007)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.006 (1.004\u0026ndash;1.008)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGlucoseAvg (mmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.77 (0.72\u0026ndash;0.83)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.11 (1.04\u0026ndash;1.18)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.003\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eC-Peptide (ng/mL)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.68 (0.63\u0026ndash;0.74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.87 (0.80\u0026ndash;0.95)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.002\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCholesterol (mmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.15 (1.07\u0026ndash;1.23)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.99 (0.88\u0026ndash;1.11)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.811\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDPP4i (No vs Yes)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.02 (0.83\u0026ndash;1.25)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.876\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.10 (0.78\u0026ndash;1.54)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.597\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGLP1-RA (No vs Yes)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.18 (1.04\u0026ndash;4.58)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.039\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.64 (0.63\u0026ndash;11.09)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.186\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSGLT2i (No vs Yes)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.17 (0.87\u0026ndash;1.58)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.289\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.29 (1.21\u0026ndash;4.32)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.011\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAGI (No vs Yes)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.57 (0.47\u0026ndash;0.70)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.67 (0.47\u0026ndash;0.94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.019\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetformin (No vs Yes)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.04 (1.66\u0026ndash;2.51)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.08 (2.19\u0026ndash;4.34)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZD (No vs Yes)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.89 (1.16\u0026ndash;3.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.011\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.24 (0.89\u0026ndash;5.64)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.086\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInsulin (No vs Yes)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.06 (0.04\u0026ndash;0.07)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.05 (0.03\u0026ndash;0.08)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eIn the multivariate multinomial logistic regression analysis (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e), independent factors associated with mild hypoglycemia included older age (OR\u0026thinsp;=\u0026thinsp;0.966, 95% CI: 0.951\u0026ndash;0.981, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), higher Charlson comorbidity index (OR\u0026thinsp;=\u0026thinsp;2.684, 95% CI: 2.121\u0026ndash;3.398, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), increased creatinine (OR\u0026thinsp;=\u0026thinsp;1.005, 95% CI: 1.003\u0026ndash;1.007, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), elevated HbA1c (OR\u0026thinsp;=\u0026thinsp;10.570, 95% CI: 7.554\u0026ndash;14.789, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), lower mean glucose levels (OR\u0026thinsp;=\u0026thinsp;0.545, 95% CI: 0.479\u0026ndash;0.619, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and insulin use (OR\u0026thinsp;=\u0026thinsp;0.205, 95% CI: 0.127\u0026ndash;0.332, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e \u003cp\u003eFor moderate-to-severe hypoglycemia, independent predictors included older age (OR\u0026thinsp;=\u0026thinsp;0.973, 95% CI: 0.954\u0026ndash;0.993, P\u0026thinsp;=\u0026thinsp;0.007), elevated Charlson comorbidity index (OR\u0026thinsp;=\u0026thinsp;2.744, 95% CI: 2.141\u0026ndash;3.517, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), increased serum creatinine (OR\u0026thinsp;=\u0026thinsp;1.005, 95% CI: 1.003\u0026ndash;1.007, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), elevated HbA1c (OR\u0026thinsp;=\u0026thinsp;10.047, 95% CI: 7.127\u0026ndash;14.163, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), lower mean glucose levels (OR\u0026thinsp;=\u0026thinsp;0.633, 95% CI: 0.554\u0026ndash;0.723, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001), and insulin use (OR\u0026thinsp;=\u0026thinsp;0.170, 95% CI: 0.086\u0026ndash;0.336, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Additionally, overweight status was a protective factor against moderate-to-severe hypoglycemia (OR\u0026thinsp;=\u0026thinsp;0.511, 95% CI: 0.302\u0026ndash;0.865, P\u0026thinsp;=\u0026thinsp;0.012), while underweight status significantly increased the risk (OR\u0026thinsp;=\u0026thinsp;3.654, 95% CI: 1.045\u0026ndash;12.781, P\u0026thinsp;=\u0026thinsp;0.043).\u003c/p\u003e \u003cp\u003eOverall, the multivariate logistic regression model demonstrated good fit (χ\u0026sup2; = 1585.430, df\u0026thinsp;=\u0026thinsp;40, P\u0026thinsp;\u0026lt;\u0026thinsp;0.001; Cox-Snell R\u0026sup2; = 0.586; Nagelkerke R\u0026sup2; = 0.703; McFadden R\u0026sup2; = 0.492).\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab3\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eMultivariate multinomial logistic regression results for mild and moderate-to-severe hypoglycemia\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"5\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMild hypoglycemia vs Normal (OR [95%CI])\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eP value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eModerate-to-Severe hypoglycemia vs Normal (OR [95%CI])\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eP value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.966 (0.951\u0026ndash;0.981)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.973 (0.954\u0026ndash;0.993)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCharlson Score\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e2.684 (2.121\u0026ndash;3.398)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e2.744 (2.141\u0026ndash;3.517)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDrug Class Count\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.788 (0.571\u0026ndash;1.086)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.146\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.791 (0.520\u0026ndash;1.202)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.272\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTG\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.857 (0.712\u0026ndash;1.033)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.105\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.904 (0.717\u0026ndash;1.139)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.391\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCreatinine\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.005 (1.003\u0026ndash;1.007)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.005 (1.003\u0026ndash;1.007)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHbA1c\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e10.570 (7.554\u0026ndash;14.789)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e10.047 (7.127\u0026ndash;14.163)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eC-Peptide\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.023 (0.945\u0026ndash;1.108)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.570\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.092 (0.994\u0026ndash;1.200)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.067\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGlucose Avg\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.545 (0.479\u0026ndash;0.619)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.633 (0.554\u0026ndash;0.723)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCholesterol\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.021 (0.894\u0026ndash;1.166)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.758\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.856 (0.728\u0026ndash;1.006)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.059\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex (Male)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.995 (0.692\u0026ndash;1.431)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.980\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.086 (0.682\u0026ndash;1.730)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.728\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSGLT2 Not Used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.953 (0.500\u0026ndash;1.817)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.883\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.809 (0.714\u0026ndash;4.580)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.211\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAGI Not Used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.680 (0.425\u0026ndash;1.089)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.109\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.680 (0.362\u0026ndash;1.277)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.230\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMetformin Not Used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.034 (0.628\u0026ndash;1.701)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.895\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.162 (0.607\u0026ndash;2.223)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.651\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTZD Not Used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.640 (0.673\u0026ndash;3.999)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.277\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e1.601 (0.475\u0026ndash;5.390)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.447\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInsulin Not Used\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.205 (0.127\u0026ndash;0.332)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.170 (0.086\u0026ndash;0.336)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI Class - Overweight\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.043 (0.699\u0026ndash;1.557)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.836\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.511 (0.302\u0026ndash;0.865)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.012\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI Class - Obese\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.254 (0.765\u0026ndash;2.057)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.370\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.747 (0.380\u0026ndash;1.470)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.399\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI Class - Underweight\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e1.964 (0.600\u0026ndash;6.427)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.265\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e3.654 (1.045\u0026ndash;12.781)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.043\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eMachine Learning Model Evaluation\u003c/p\u003e \u003cp\u003eThe predictive performances of multinomial logistic regression, XGBoost, and random forest (RF) models were evaluated using receiver operating characteristic (ROC) curves constructed through a One-vs-Rest approach. ROC curves for each model are illustrated individually in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e,Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e,Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e, and a direct comparative analysis among the three models is presented in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e.\u003c/p\u003e \u003cp\u003eThe multinomial logistic regression model yielded area under the curve (AUC) values of 0.938 for predicting normoglycemia, 0.889 for mild hypoglycemia, and 0.754 for moderate-to-severe hypoglycemia. The XGBoost model demonstrated notably higher AUCs of 0.991, 0.980, and 0.943, respectively. Similarly, the RF model achieved AUCs of 0.991, 0.983, and 0.952 for the three respective severity categories.\u003c/p\u003e \u003cp\u003eComparative ROC analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e) clearly indicated superior discriminative performance of both XGBoost and RF models across all hypoglycemia severity categories compared to logistic regression.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eModel Performance Comparison\u003c/p\u003e \u003cp\u003eTo comprehensively evaluate predictive performances, we compared the overall accuracy, Kappa coefficient, and macro-average AUC of the three models (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). The RF model achieved the highest accuracy (93.3%) and Kappa coefficient (0.873), followed closely by XGBoost (accuracy: 92.6%, Kappa: 0.860). Logistic regression demonstrated relatively lower performance, with an accuracy of 83.8% and Kappa of 0.685. Consistently, the macro-average AUC values for RF, XGBoost, and logistic regression were 0.960, 0.955, and 0.788, respectively, further emphasizing the superior predictive capability of the tree-based machine learning algorithms.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab4\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 4\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003ePerformance comparison of three models: Accuracy, AUC, and Kappa\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eModel\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAccuracy\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eKappa\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eMacro AUC\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eLogistic Regression\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.838\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.6845\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.7878\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eXGBoost\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.9255\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8603\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.9553\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e\u003cb\u003eRandom Forest\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e \u003cp\u003e0.9330\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.8729\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.9596\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003eFeature Importance Analysis\u003c/p\u003e \u003cp\u003eFeature importance analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e) identified mean blood glucose level and HbA1c as the two most critical predictors in both XGBoost and RF models. Subsequent important predictors in the XGBoost model included creatinine, C-peptide, and triglycerides (TG). In the RF model, following mean glucose and HbA1c, important predictors included C-peptide, insulin usage, and the Charlson comorbidity index.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn the present study, we developed and systematically compared three predictive models\u0026mdash;multinomial logistic regression, XGBoost, and random forest (RF)\u0026mdash;to identify key predictors of hypoglycemia severity among hospitalized patients with diabetes. Our results demonstrated that both machine learning models (XGBoost and RF) exhibited superior discriminative performance compared to traditional logistic regression, especially for detecting moderate-to-severe hypoglycemia, this has some modeling studies with the same results\u003csup\u003e[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]\u003c/sup\u003e. We utilized the One-vs-Rest strategy, a standard and effective approach for generating ROC curves in multiclass classification problems, allowing robust evaluation and comparison of each model's predictive performance across hypoglycemia categories.\u003c/p\u003e \u003cp\u003eThe dataset used to construct our predictive models comprised common clinical and laboratory variables routinely available for hospitalized patients with type 2 diabetes mellitus (T2DM). The logistic regression model identified seven statistically significant predictors (P\u0026thinsp;\u0026lt;\u0026thinsp;0.05), while both XGBoost and RF models highlighted the top 10 most influential variables. Notably, six predictors\u0026mdash;age, Charlson Comorbidity Index (CCI), serum creatinine, HbA1c, mean glucose levels, and insulin use\u0026mdash;emerged consistently as significant factors across all three models. A number of previous modeling studies and factor analysis studies have similarly given the same conclusions for one or more of these factors\u003csup\u003e[\u003cspan additionalcitationids=\"CR9\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]\u003c/sup\u003e. While previous studies have individually confirmed the significance of one or several of these factors, our research uniquely integrated multiple models for comprehensive comparison, offering a broader perspective than single-model studies previously reported.\u003c/p\u003e \u003cp\u003eA key advantage of our study lies in its reliance on routinely collected clinical data. This facilitates rapid assessment of hypoglycemia risk without additional specialized testing, thus granting medical staff valuable time to implement preventive strategies and appropriate treatments.\u003c/p\u003e \u003cp\u003eRegarding predictive variables, HbA1c has consistently been identified as a crucial factor in hypoglycemia research. Prior studies have reported that both excessively high HbA1c levels (\u0026gt;\u0026thinsp;9%) and overly stringent glycemic control (HbA1c\u0026thinsp;\u0026lt;\u0026thinsp;7%) are associated with increased hypoglycemia risk\u003csup\u003e[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]\u003c/sup\u003e. Additionally, insulin use within these HbA1c ranges further exacerbates hypoglycemia risk. Both our univariate and multivariate analyses supported these findings, reinforcing that elevated HbA1c levels and insulin therapy are critical risk factors for hypoglycemia in diabetic patients.\u003c/p\u003e \u003cp\u003eIn recent years, growing attention has focused on the association between the Charlson Comorbidity Index (CCI) and hypoglycemia. Most prior studies investigating comorbidities typically examined only one or a few components of the CCI, such as cardiovascular disease, renal impairment, or malignancy\u003csup\u003e[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]\u003c/sup\u003e. These comorbid conditions have been associated with glucose instability and heightened risk of adverse glycemic events. Patients with multiple chronic diseases often experience altered drug metabolism, polypharmacy, and malnutrition, potentially interfering with glucose regulation and insulin sensitivity. Incorporating the CCI into all three predictive models in our study enhances its utility as a comprehensive clinical indicator. Unlike individual diagnoses, the CCI provides an aggregated measure of overall disease burden, capturing complex interactions between comorbidities and thereby improving the generalizability and interpretability of the predictive models. Integrating the CCI into predictive assessments may assist clinicians in accurately stratifying risk, particularly among elderly hospitalized patients or those with multiple chronic diseases, who inherently have higher susceptibility to hypoglycemic episodes\u003csup\u003e[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]\u003c/sup\u003e. Early identification of high-CCI patients could facilitate personalized monitoring, medication adjustments, and tailored nutritional plans, thereby mitigating hypoglycemia risk.\u003c/p\u003e \u003cp\u003eLastly, we conducted feature importance analyses for the top ten variables in the XGBoost and RF models. Both models consistently identified glycemic control and glucose variability as key predictors. Additionally, XGBoost emphasized basal metabolic parameters (e.g., creatinine and C-peptide), whereas the RF model placed greater emphasis on medication use. These findings highlight nuanced differences between machine learning approaches, underscoring their ability to provide targeted clinical insights into managing and preventing hypoglycemic events in hospitalized diabetic patients.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eIn this study, we developed and compared three prediction models - multinomial logistic regression, XGBoost, and random forest - to identify factors associated with hypoglycemia severity in hospitalized diabetic patients. Our results showed that both machine learning models outperformed the traditional logistic regression model in terms of predictive performance, with higher accuracy and AUC values in all hypoglycemia categories. This time, the three models were designed with all commonly used clinical indicators, without the need for special tests, to prevent hypoglycemia as early as possible in type 2 diabetic patients and to avoid the adverse consequences caused by hypoglycemia. In a comparison of the three models and the focus of the models, the six variables of age, Charlson comorbidity index, creatinine, glycosylated hemoglobin, mean blood glucose level, and insulin use were consistently identified as core predictors in almost all models \u003csup\u003e[\u003cspan additionalcitationids=\"CR15\" citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]\u003c/sup\u003e, This further enhances their clinical relevance。Future studies should focus on the prospective validation and practical application of these models to assess their clinical utility and impact on patient safety.\u003c/p\u003e \u003cp\u003eLimitations\u003c/p\u003e \u003cp\u003eThe model was designed to use only common clinical indicators, and because hypoglycaemia occurs in a variety of ways, even if the model had a higher degree of accuracy, it would still not be able to fully diagnose all types of hypoglycaemia. As this study was based on single-centre data, further validation of the predictive effect of the model is needed in the future. Conflict of interest statement\u003c/p\u003e \u003cp\u003eThe authors declare that they have no financial conflict of interest with regard to the content of this report.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eThe study titled \u0026ldquo;Machine Learning-Based Prediction of Hypoglycemia Severity in Hospitalized Diabetic Patients\u0026rdquo; was reviewed and approved by the Ethics Committee of The Affiliated Hospital of Qingdao University. All procedures involving human participants were conducted in accordance with institutional guidelines and approved protocols.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eHongjian Jia contributed to study design, data preprocessing, modeling, and manuscript writing and is the first author.Jietao Zhang supervised the project, interpreted clinical significance, and is the corresponding author.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eONG K L, STAFFORD L K, MCLAUGHLIN, S. A. et al. Global, regional, and national burden of diabetes from 1990 to 2021, with projections of prevalence to 2050: a systematic analysis for the Global Burden of Disease Study 2021[J]. \u003cem\u003eLancet\u003c/em\u003e \u003cb\u003e402\u003c/b\u003e (10397), 203\u0026ndash;234. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/S0140-6736(23)01301-6\u003c/span\u003e\u003cspan address=\"10.1016/S0140-6736(23)01301-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNUGRAHA A R A, Y. U. N. I. R. E. et al. Risk factors of severe hypoglycemia among patients with type 2 diabetes mellitus in outpatient clinic of tertiary hospital in Indonesia[J]. \u003cem\u003eSci. Rep.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e (1), 16259. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41598-023-43459-2\u003c/span\u003e\u003cspan address=\"10.1038/s41598-023-43459-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2023).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBONAVENTURA, A., MONTECUCCO, F. \u0026amp; DALLEGRI, F. Update on strategies limiting iatrogenic hypoglycemia[J/OL]. (2015-09-01)[2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://ec.bioscientifica.com/view/journals/ec/4/3/R37.xml\u003c/span\u003e\u003cspan address=\"https://ec.bioscientifica.com/view/journals/ec/4/3/R37.xml\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (2015). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1530/EC-15-0044\u003c/span\u003e\u003cspan address=\"10.1530/EC-15-0044\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnonymous Predictors of hypoglycemia in hospitalized patients with diabetes mellitus | Internal and Emergency Medicine[EB/OL]. [2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://link.springer.com/article/10.1007/s11739-018-1787-0\u003c/span\u003e\u003cspan address=\"https://link.springer.com/article/10.1007/s11739-018-1787-0\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnonymous. Risk factors for hypoglycemia in patients with type 2 diabetes, hospitalized in internal medicine wards: Findings from the FADOI-DIAMOND study - Diabetes Research and Clinical Practice[EB/OL]. [2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.diabetesresearchclinicalpractice.com/article/S0168-8227(16)00069-3/abstract\u003c/span\u003e\u003cspan address=\"https://www.diabetesresearchclinicalpractice.com/article/S0168-8227(16)00069-3/abstract\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnonymous Rates and predictors of hypoglycaemia in 27 585 people from 24 countries with insulin-treated type 1 and type 2 diabetes: the global HAT study - Khunti \u0026ndash;\u0026thinsp;2016 - Diabetes, Obesity and Metabolism - Wiley Online Library[EB/OL]. [2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://dom-pubs.onlinelibrary.wiley.com/doi/full/10.1111/dom.12689\u003c/span\u003e\u003cspan address=\"https://dom-pubs.onlinelibrary.wiley.com/doi/full/10.1111/dom.12689\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnonymous Interdisciplinary Nursing Research[EB/OL]. [2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://journals.lww.com/inr/fulltext/2025/03000/machine_learning_based_prediction_model_for.4.aspx\u003c/span\u003e\u003cspan address=\"https://journals.lww.com/inr/fulltext/2025/03000/machine_learning_based_prediction_model_for.4.aspx\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnonymous The Risk Factors of Severe Hypoglycemia in Older Patients with Dementia and Type 2 Diabetes Mellitus - PMC[EB/OL]. [2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://pmc.ncbi.nlm.nih.gov/articles/PMC8779381/?utm_source=chatgpt.com#sec5-jpm-12-00067\u003c/span\u003e\u003cspan address=\"https://pmc.ncbi.nlm.nih.gov/articles/PMC8779381/?utm_source=chatgpt.com#sec5-jpm-12-00067\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnonymous Hypoglycemia in patients with type 2 diabetes mellitus during hospitalization: associated factors and prognostic value | Diabetology \u0026amp; Metabolic Syndrome | Full Text[EB/OL]. [2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://dmsjournal.biomedcentral.com/articles/\u003c/span\u003e\u003cspan address=\"https://dmsjournal.biomedcentral.com/articles/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s13098-023-01212-9?utm_source=chatgpt.com#Sec8\u003c/span\u003e\u003cspan address=\"10.1186/s13098-023-01212-9?utm_source=chatgpt.com#Sec8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAKIROV, A. et al. Predictors of hypoglycemia in hospitalized patients with diabetes mellitus[J]. \u003cem\u003eIntern. Emerg. Med.\u003c/em\u003e \u003cb\u003e13\u003c/b\u003e (3), 343\u0026ndash;350. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1007/s11739-018-1787-0\u003c/span\u003e\u003cspan address=\"10.1007/s11739-018-1787-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2018).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLIPSKA K J et al. HbA1c and Risk of Severe Hypoglycemia in Type 2 Diabetes[J]. \u003cem\u003eDiabetes Care\u003c/em\u003e. \u003cb\u003e36\u003c/b\u003e (11), 3535\u0026ndash;3542. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.2337/dc13-0610\u003c/span\u003e\u003cspan address=\"10.2337/dc13-0610\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e (2013).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnonymous. The risk factors of inpatient hypoglycemia: A systematic review - PMC[EB/OL]. [2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://pmc.ncbi.nlm.nih.gov/articles/PMC7218453/?utm_source=chatgpt.com#sec3\u003c/span\u003e\u003cspan address=\"https://pmc.ncbi.nlm.nih.gov/articles/PMC7218453/?utm_source=chatgpt.com#sec3\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnonymous Prevalence and predictors of hypoglycemia in older outpatients with type 2 diabetes mellitus | PLOS One[EB/OL]. [2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0309618\u0026amp;utm_source=chatgpt.com#sec016\u003c/span\u003e\u003cspan address=\"https://journals.plos.org/plosone/article?id=10.1371%2Fjournal.pone.0309618\u0026amp;utm_source=chatgpt.com#sec016\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnonymous. Predicting hypoglycemia in elderly inpatients with type 2 diabetes: the ADOCHBIU model - PMC[EB/OL]. [2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://pmc.ncbi.nlm.nih.gov/articles/PMC11602273/?utm_source=chatgpt.com\u003c/span\u003e\u003cspan address=\"https://pmc.ncbi.nlm.nih.gov/articles/PMC11602273/?utm_source=chatgpt.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnonymous. Enhancing severe hypoglycemia prediction in type 2 diabetes mellitus through multi-view co-training machine learning model for imbalanced dataset | Scientific Reports[EB/OL]. [2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.nature.com/articles/s41598-024-69844-z?utm_source=chatgpt.com\u003c/span\u003e\u003cspan address=\"https://www.nature.com/articles/s41598-024-69844-z?utm_source=chatgpt.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAnonymous. Predicting the Risk of Inpatient Hypoglycemia With Machine Learning Using Electronic Health Records. | Diabetes Care | American Diabetes Association[EB/OL]. [2025-05-01]. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://diabetesjournals.org/care/article/43/7/1504/35530/Predicting-the-Risk-of-Inpatient-Hypoglycemia-With?utm_source=chatgpt.com\u003c/span\u003e\u003cspan address=\"https://diabetesjournals.org/care/article/43/7/1504/35530/Predicting-the-Risk-of-Inpatient-Hypoglycemia-With?utm_source=chatgpt.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Type 2 diabetes mellitus, Hypoglycemia, Risk prediction, Machine learning","lastPublishedDoi":"10.21203/rs.3.rs-6619286/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6619286/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eObjective: \u003c/strong\u003eTo identify risk factors for hypoglycemia in hospitalized patients with type 2 diabetes mellitus (T2DM) and develop predictive models for hypoglycemia severity based on machine learning algorithms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods: \u003c/strong\u003eAdult non-pregnant hospitalized patients diagnosed with T2DM were retrospectively enrolled from the electronic medical record system of the Affiliated Hospital of Qingdao University. Patients were categorized into hypoglycemia groups (mild, moderate-to-severe) or a non-hypoglycemia group based on inpatient venous plasma glucose levels. After data preprocessing, univariate and multivariate analyses were conducted to identify significant predictors. Three predictive models (XGBoost, Random Forest [RF], and Logistic Regression) were subsequently constructed and validated to evaluate their predictive performances.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults: \u003c/strong\u003eFrom an initial cohort of 8,947 patients, 1,798 patients were included after data screening. Among the evaluated models, the RF model demonstrated the highest predictive accuracy (93.3%) and Kappa coefficient (0.873), followed by XGBoost (accuracy: 92.6%, Kappa: 0.860). Logistic regression exhibited comparatively lower performance (accuracy: 83.8%, Kappa: 0.685). The macro-average area under the ROC curve (AUC) values for RF, XGBoost, and logistic regression were 0.960, 0.955, and 0.788, respectively, highlighting the superior discriminative capability of the RF model. While both XGBoost and RF models identified glycemic control metrics and glucose variability as core predictors for hypoglycemia, the RF model additionally emphasized medication usage, whereas XGBoost prioritized basal metabolic parameters.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions: \u003c/strong\u003eThe RF model outperformed XGBoost and conventional logistic regression in predicting hypoglycemia severity among hospitalized T2DM patients. The results emphasize the importance of closely monitoring glucose levels and glucose variability during diabetes management to prevent hypoglycemia. The developed model provides a foundation for implementing preventive strategies to reduce hypoglycemia occurrence in hospitalized patients with T2DM.\u003c/p\u003e","manuscriptTitle":"Machine Learning-Based Prediction of Hypoglycemia Severity in Hospitalized Diabetic Patients","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-16 12:18:54","doi":"10.21203/rs.3.rs-6619286/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"56439e66-d1ba-49a4-a3c9-b4c3c1ce6f74","owner":[],"postedDate":"May 16th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":48247417,"name":"Health sciences/Diseases"},{"id":48247418,"name":"Health sciences/Endocrinology"},{"id":48247419,"name":"Health sciences/Medical research"}],"tags":[],"updatedAt":"2025-05-23T10:38:40+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-16 12:18:54","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6619286","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6619286","identity":"rs-6619286","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}