A study on the development of a machine learning-based clinical prediction model for bone health management in breast cancer patients

A study on the development of a machine learning-based clinical prediction model for bone health management in breast cancer 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 Research Article A study on the development of a machine learning-based clinical prediction model for bone health management in breast cancer patients Meiling Chu, Xiaowen Lai, Guanghua Yang, Lifang Song This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6455887/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 17 You are reading this latest preprint version Abstract Background Breast cancer is among the most prevalent malignant neoplasms afflicting women globally, with its treatment and disease itself exerting a substantial influence on bone health. Given the prolonged survival rates of breast cancer patients, the management of bone health issues has become a critical component of comprehensive cancer care. The objective of this study was to develop clinical predictive models using machine learning methods and apply these models to the management of bone health in breast cancer patients. Methods This was a multicenter retrospective cohort study. We included 139 breast cancer patients who were diagnosed from March 2022 to December 2024 in three hospitals in China. We developed predictive models with optimal features by using algorithms such as random forest (RF), K nearest neighbors (KNN), support vector machines (SVM), and extreme gradient boosting (XGB) and determined and assessed the machine learning algorithm with the highest accuracy rate for breast cancer-related bone loss on the basis of the area under curve (AUC) of the subjects. Results A total of 139 study participants were included in this study, including 78 patients with osteopenia (including osteoporosis, T≤-1.0) and 61 patients with normal bone mass (T>-1.0). Specific indicators of bone loss were identified, and seven models were constructed: logistic regression (LR), the bagging tree algorithm (BT), RF, adaptive boosting (AdaBoost), XGBoost, KNN, and SVM, among which the AdaBoost prediction model performed the best in predicting breast cancer-related bone loss, with the best performance AUC of 0.995. The model can be generated into a publicly accessible applet. Conclusion The data was incorporated into seven machine learning algorithms and models were constructed, which were then compared to arrive at the optimal model.The AdaBoost model was selected as the final model, which can predict breast cancer-related bone loss by collecting patient information through a simple question-and-answer session and importing laboratory blood tests and reports of examinations such as bone mineral density; this model is expected to guide the management of clinical bone health in patients with breast cancer and improve the prognosis of patients. Trial registration Ethics Approval and Consent to Participate This study was approved by the Institutional Review Board of the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine. Approval number: ChiCTR2200057785(17th March 2022). Breast cancer Bone loss Risk prediction model Learning algorithm Figures Figure 1 Figure 2 Introduction According to the Global Cancer Statistics 2022 published by the International Agency for Research on Cancer (IARC) of the World Health Organization, breast cancer has become the malignant tumor with the highest incidence in women worldwide[ 1 ]. Studies have shown that the global incidence of breast cancer will increase every year from 2020 to 2050, and 4,781,849 cases of breast cancer and 1,503,694 deaths from breast cancer are expected to occur worldwide in 2050[ 2 ]. Notably, bone loss caused by age, disease, treatment, hormone level changes, and other factors during the diagnosis and treatment of breast cancer patients has become a concomitant symptom of the high incidence of breast cancer[ 3 ], which affects the quality of life of patients and their recovery from the disease.In 2023, the China Oncology Healthcare Management Conference promulgated the inaugural set of guidelines for standardized management of the entire cycle of bone health in breast cancer. Bone safety management for breast cancer patients refers to the early assessment, prevention, diagnosis, treatment, and rehabilitation management of bone loss, osteoporosis, and fracture risk events during disease treatment and follow-up[ 3 ]. Early bone health management for patients with breast cancer includes the prevention and treatment of bone loss caused by tumors and the prevention of bone metastasis. The Chinese Anti-Cancer Society Breast Cancer Diagnostic and Treatment Guidelines and Specifications (Version 2024) suggest making staging diagnoses and corresponding treatment plans for breast cancer staging criteria on the basis of clinical examination and surgical pathology results[ 4 ]. The musculoskeletal-related side effects of endocrine therapy, radiotherapy, chemotherapy, and other posttreatments for breast cancer seriously affect patients' quality of life and medication compliance. The incidence of abnormal bone mass in breast cancer patients in China reaches 77.7%, and the incidence of osteoporosis is 30.5%[ 5 ]. Research has shown that the incidence of bone metastasis in advanced stages of breast cancer is 65%~75%[ 4 ]. The essence of Chinese medicine lies in holistic concepts and evidence-based treatment. Chinese medicine emphasizes that the human body is an organic whole and that all parts are interrelated and affect each other. Through inspection, auscultation-olfaction, inquiry, and palpation, the four diagnostic methods are combined to analyze the cause and mechanism of the disease according to the patient's constitution to formulate a personalized treatment plan, which focuses not only on the alleviation of symptoms but also on eliminating the causes of the disease and conditioning the constitution. Therefore, a medical support system that combines Chinese and Western medicine is increasingly favored for the local recurrence and metastasis of breast cancer and its treatment. Currently, Fatima et al.[ 6 ] have conducted a comprehensive comparative analysis of machine learning techniques for breast cancer prediction, offering detailed insights into the algorithms employed by various machine learning approaches in this domain. However, developing efficient methods for breast cancer-related bone loss prediction with both iterative learning capabilities and interpretability to continuously improve model efficacy is challenging. We aim to design and develop an applet for clinical practice that facilitates an intelligent service system for health counseling, intelligent diagnosis, assisted diagnosis, and treatment decision-making for the bone health management of patients with breast cancer. To play an active role in the management of bone safety in the whole management and all-round management chain of breast cancer and to be widely used in clinical practice and welfare society. Methods Study population This multicenter retrospective study included breast cancer patients from March 2022 to December 2024 from three tertiary-level hospitals in China: the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine, the Longhua Hospital affiliated with Shanghai University of Traditional Chinese Medicine, and the Shengjing Hospital Affiliated with China Medical University. Inclusion and Exclusion criteria The inclusion criteria were as follows: (1) aged between 18 and 80 years; (2) KPS score ≥ 60; and (3) diagnosed with breast cancer by pathology. The exclusion criteria were as follows: (1) bone mineral diseases, such as deformational osteitis, osteogenesis imperfecta, osteochondrosis, etc.; (2) diseases affecting bone metabolism, such as Cushing's syndrome, rheumatoid or rheumatoid arthritis, etc.; (3) hormone replacement therapy used in the past 6 months, such as glucocorticoids, parathyroid hormone, etc.; (4) surgical treatment for limb joints carried out in the past 6 months; (5) severe cardiovascular, liver, renal or hematopoietic systems and other serious primary diseases; (6) psychiatric patients with cognitive loss or mental abnormality; and (7) insufficient clinical information in the medical file. Data collection Before initiating data collection, all relevant staff involved were trained on the data extraction form. The medical records of eligible patients were meticulously reviewed. Clinical and laboratory data (including demographic characteristics, TCM symptom scales, routine blood tests, paraffin pathology reports of breast cancer, and T values of bone density measured by dual-energy X-ray absorptiometry) were systematically extracted into standardized forms for this project. To ensure the relevance and consistency of the data, 30% of the randomly extracted data were evaluated and validated by the principal. Units of measurement were standardized across all laboratories, taking into account differences in units across participating centers. Extreme outliers, either significantly large or significantly small values, were flagged for review. These outliers were reevaluated by the program principal or designated attending physician to determine the validity of the data and to rule out data entry errors. Overview of machine learning models We developed and compared seven models, including LR, the Bagging Tree algorithm, RF, AdaBoost, XGBoost, KNN, and SVM, for predicting the risk of bone loss and bone metastasis in patients with breast cancer. Logistic regression is a method widely used for binary outcomes, especially in the social sciences, epidemiology, and machine learning[ 7 ]. The Bagging Tree algorithm belongs to a wider range of integrated methods that improve the stability and accuracy of tree-based methods used in nonparametric regression, mainly through voting or model averaging[ 8 ]. Random forest is a further improvement in classification and regression tree models that integrates multiple decision trees to better identify meaningful interactions and nonlinear effects in predictor variables[ 9 ]. AdaBoost is an integrated learning process that employs an iterative approach to improve poor-quality classifiers by learning from errors[ 10 ]. XGBoost is a scalable machine for tree boosting learning systems for dealing with sparse data, which increases the speed of learning and thus speeds up model exploration [ 11 ]. KNN is a supervised machine learning algorithm applied for classification purposes that can be trained to predict the classification of unknown data by training the features and labels of the known data, and it is now widely used for disease prediction[ 12 ]. SVM is a discriminative classifier characterized by an optimal hyperplane to maximize the interval between different classes[ 13 ]. Primary objective The primary objective of this study was to construct a machine learning model for rapid adjuvant diagnosis and treatment for the bone health management of breast cancer patients. Statistical analysis Patient data are presented as categorical or continuous variables. The Kolmogorov‒Smirnov test was used to assess whether the data were normally distributed. For normally distributed continuous variables, the data are presented as the means ± standard deviations and were compared via t tests. If continuous variables did not follow a normal distribution, the Mann‒Whitney U test was used, and the results are expressed as the median (interquartile spacing). Categorical data are expressed as numbers and frequencies and were compared via the chi-square test or Fisher's exact test. All the statistical tests were two-sided, and p < 0.05 indicated statistical significance. SPSS software version 25.0 was used in this study. Results This study included a total of 66 candidate predictors, including basic patient information, disease diagnosis and treatment,the Fracture Risk Assessment Tool (FRAX) and the Western Ontario McMaster University Osteoarthritis Index Scale (WOMA), involving 25 characteristic variables, as well as 25 blood index testing items and 16 TCM symptom elements.The 66 variables from the training cohort were incorporated into 7 machine algorithms, namely, the LR, Bagging Tree, RF, AdaBoost, XGBoost, KNN, and SVM algorithms, to generate a prediction model for predicting bone loss in patients with breast cancer, and the prediction model was used to provide a quantitative description of the overall relationship between breast cancer-related bone loss and all 66 features. The features were characterized using the Shapley significance analysis, and 10 highly sensitive features were screened out. In this process, the "contribution" of each feature to the model is calculated on the basis of the marginal contribution of each feature and the number of occurrences of the feature, and the optimal algorithm is selected after the performance differences between different machine learning algorithms are compared. The performance of the model is evaluated usingvia two methods: the ROC curve and the confusion matrix. The confusion matrix shows the number of correctly predicted samples and the number of incorrectly predicted samples in each category in the form of a matrix (Table 1 ). Table 1 Confusion matrix Condition positive Condition positive Test outcome positive True Positive (TP) False Positive (FP) Test outcome negative False Negative (FN) True Negative (TN) Accuracy: represents the proportion of the correct sample size to the total sample size. Role of the funding source The funder of this study had no role in the research design, data collection, data analysis, data interpretation, or report writing. 1 Patient characteristics Table 2 Baseline characteristics of the cohort Variables Normal group (n = 61) Osteopenia group (n = 78) p Value (2-sided) Age 47.49 ± 8.57 49.87 ± 9.038 0.118 BMI (kg/m 2) 23.94 ± 3.69 23.14 ± 3.128 0.157 Education High school and below 32(52.5%) 41(52.6%) 0.99 More than university 29(47.5%) 37(47.4%) Age at menarche 13.77 ± 1.383 14.09 ± 1.321 0.106 Menopause No 39(63.9%) 41(52.6%) 0.178 Yes 22(36.1%) 37(47.4%) Number of pregnancies No 6(9.8%) 1(1.3%) 0.072 Once or twice 40(65.6%) 57(73.1%) More than twice 15(24.6%) 20(25.6%) Number of production times No 6(9.8%) 3(3.8%) 0.339 Once 41(67.2%) 58(74.4%) Twice and more 14(23%) 17(21.8%) Duration of breast cancer (months) 17.11 ± 10.767 26.54 ± 15.965 <0.001 Family history of cancer 18(29.5%) 24(30.8%) 0.872 Lymph node metastasis 27(44.3%) 32(41.0%) 0.702 Targeted therapy 9(14.8%) 17(21.8%) 0.291 Radiotherapy 36(59.0%) 47(60.3%) 0.882 Chemotherapy 27(44.3%) 39(50%) 0.501 Drink coffee 13(21.3%) 13(16.7%) 0.486 Exercise 25(41%) 36(46.2%) 0.542 History of fractures 6(9.8%) 10(12.8%) 0.584 History of parental hip fracture 7(11.5%) 17(21.8%) 0.11 L1-L4 BMD 0.177 ± 0.910 -1.468 ± 0.842 <0.001 Femur BMD -0.151 ± 0.569 -1.124 ± 0.828 <0.001 FRAX Major Osteoporotic Fracture 2.303 ± 1.741 3.705 ± 3.255 <0.001 Hip Fracture 0.336 ± 1.064 0.851 ± 1.63 <0.001 WOMAC 26.21 ± 15.80 30.13 ± 15.485 0.3 Table 3 Analysis of blood indicators in the study population Variables Normal group (n = 61) Osteopenia group (n = 78) p Value (2-sided) Growth hormone (ug/L) 0.879 ± 1.125 1.015 ± 1.329 0.551 Estradiol (pg/mL) 23.572 ± 97.177 20.913 ± 85.362 0.273 Follicle-stimulating hormone (IU/L) 32.404 ± 33.001 34.472 ± 31.274 0.257 Luteinizing hormone (IU/L) 11.622 ± 14.783 12.56 ± 15.194 0.686 Prolactin (ng/mL) 190.966 ± 138.584 149.568 ± 96.881 0.135 Progesterone (ng/mL) 0.625 ± 0.349 0.781 ± 0.555 0.275 Testosterone (nmol/L) 1.142 ± 1.233 1.245 ± 2.644 0.569 25-Hydroxyvitamin D (nmol/L) 64.83 ± 24.39 71.388 ± 22.233 0.10 Osteocalcin (ng/mL) 26.039 ± 10.665 23.641 ± 9.483 0.14 β-Isomerized C-telopeptide (pg/mL) 639.700 ± 250.535 566.352 ± 273.333 0.106 Alanine aminotransferase (U/L) 21.802 ± 15.195 21.923 ± 14.979 0.963 Aspartate aminotransferase (U/L) 22.367 ± 11.842 21.617 ± 7.124 0.553 Serum creatinine (µmol/L) 52.362 ± 13.343 54.466 ± 10.673 0.32 Blood urea nitrogen (mmol/L) 5.948 ± 8.560 4.984 ± 1.184 0.609 γ-Glutamyl transpeptidase (U/L) 28.104 ± 19.926 25.715 ± 17.79 0.724 Alkaline phosphatase (U/L) 79.590 ± 26.444 87.997 ± 35.428 0.198 Total bilirubin (µmol/L) 12.421 ± 5.688 13.41 ± 7.188 0.158 Calcium (mmol/L) 2.389 ± 0.128 2.399 ± 0.136 0.587 Total protein (g/L) 72.038 ± 6.523 73.677 ± 3.955 0.392 White blood cell count (×10 9 /L) 5.485 ± 1.738 4.909 ± 1.03 0.049 Neutrophil count (×10 9 /L) 3.318 ± 1.572 2.794 ± 0.784 0.06 Lymphocyte count (×10 9 /L) 1.775 ± 0.556 2.094 ± 3.317 0.914 Red blood cell count (×10 12 /L) 4.399 ± 0.443 4.477 ± 0.339 0.502 Hemoglobin (g/L) 132.541 ± 13.581 134.718 ± 8.197 0.74 Platelet count (×10 9 /L) 213.72 ± 54.667 205.33 ± 43.188 0.314 Table 4 Analysis of the Chinese medicine symptoms of the study subjects Variables Normal group (n = 61) Osteopenia group (n = 78) p Value (2-sided) Pain 56(91.8%) 75(96.2%) 0.299 Lumbar-knee soreness with weakness 53(86.9%) 65(83.3%) 0.562 Gait disturbance 39(63.9%) 53(67.9%) 0.62 Cramping in the lower extremities 19(31.1%) 34(43.6%) 0.134 Morning stiffness; numbness in the finger joints 48(78.7%) 71(91%) 0.040 Cold intolerance and chilly extremities 34(55.7%) 51(65.4%) 0.247 Increased nocturia 46(75.4%) 56(71.8%) 0.632 Loose stool 26(42.6%) 42(53.8%) 0.189 Shortness of breath and fatigue 47(77%) 70(89.7%) 0.042 Hot flashes and sweating 51(83.6%) 68(87.2%) 0.551 Tidal fever and night sweats 45(73.8%) 57(73.1%) 0.927 Dry throat and mouth 47(77%) 60(76.9%) 0.986 Dizziness and blurred vision 40(65.6%) 55(70.5%) 0.534 Tinnitus 31(50.8%) 39(50%) 0.924 Memory decline 54(88.5%) 77(98.7%) 0.01 Insomnia 48(78.7%) 62(79.5%) 0.908 A total of 139 study participants were categorized into 61 individuals in the normal bone mass group (T > 1.0) and 78 individuals in the reduced bone mass group (including those with osteoporosis) (T ≤ 1.0) on the basis of the T value of BMD. The baseline characteristics of the study sample are shown in Table 2 . The duration of breast cancer was longer in the reduced bone mass group than in the normal bone mass group, with means of 17.11 ± 10.767 and 26.54 ± 15.965, respectively. The overall probability of osteoporosis and the probability of hip fracture were statistically significant in both groups, as calculated by the Fracture Risk Assessment Tool (both p values were less than 0.001). The white blood cell count was significantly different ( p = 0.049) between the two groups (Table 3 ). A comparison of the TCM evidence scale between the two groups also revealed favorable results, with the incidence of morning stiffness, numbness of the knuckles, shortness of breath, and memory loss in the normal bone mass group (78.7%, 77%, and 88.5%, respectively) and the reduced bone mass group (91%, 89.7%, and 98.7%, respectively) (Table 4 ). 2 Performance of each predictive model in identifying breast cancer-associated bone loss Sixty-six variables from the training cohort were integrated into the ML algorithm to create predictive models, and 10 variables with high significance were evaluated. The final RF model included L1‒L4 BMD, femur BMD, duration of breast cancer, MOF, HF, TBil, β-CTx, LH, TP, and WOMAC scores (Fig. 1 A), which had poorer accuracy than the other models in the test set of 0.810 (Fig. 2 B). The variables included in the LR model included L1-L4 BMD, femur BMD, dizziness and blurred vision, shortness of breath and fatigue, AST, OC, insomnia, increased nocturia, SCr, and testosterone (Fig. 1 B), and the model's accuracies in the training and test sets were 1.00 and 0.857, respectively (Fig. 2 B). The SVM model included L1–L4 BMD, insomnia, dizziness and blurred vision, the WOMAC score, femur BMD, AST, OC, shortness of breath and fatigue, T, ALP (Fig. 1 C), and the model had an accuracy of 0.833 in the test set (Fig. 2 B). The XGBoost model included L1–L4 BMD, femur BMD, Ca, OC, radiotherapy, memory decline, ALT, education, RBC, and TBil (Fig. 1 D), which performed well on both the training and test sets, with accuracies of 1.00 and 0.929, respectively (Fig. 2 B). The Bagging Tree model included L1‒L4 BMD, femur BMD, SCr, Ca, WOMAC score, γ-GT, HGH, ALT, β-CTx, and OC (Fig. 1 E). The accuracy of the bagging tree model accuracy was 0.952 (Fig. 2 B). KNN included L1-L4 BMD, femur BMD, FSH, shortness of breath and fatigue, PRL, BUN, education, age at menarche, duration of breast cancer, and OC (Fig. 1 F), and the training set accuracy of this model showed poor performance at 0.887 (Fig. 2 B). The AdaBoost model included OC, radiotherapy, L1-L4 BMD,RBC, TP, 25(OH)D, Ca, memory decline, femur BMD and dizziness and blurred vision (Fig. 1 G), and the AdaBoost model had the best performance in the test set, with an accuracy of 0.952 (Fig. 2 B). The AUC of the RF, LR, Bagging Tree, AdaBoost, XGBoost, KNN, and SVM models were 0.957, 0.907, 0.963, 0.995, 0.991, 0.860, and 0.868, respectively (Fig. 2 A), and the best AUC was obtained by AdaBoost. Discussion In this study, we developed and validated a model based on a machine learning algorithm to predict breast cancer-related bone loss. The AdaBoost model was selected as the final model, which had an AUC of 0.995 and an accuracy of 0.952. These findings demonstrated the accuracy and stability of the model in predicting bone loss in patients with breast cancer. Our study suggests that the AdaBoost model has the potential to identify the degree of risk of developing osteoporosis in breast cancer patients, which can help clinicians grasp the risk of bone loss occurring during the management of breast cancer for further examination and assist in treatment decisions. Owing to the emphasis on breast cancer screening and advances in medical technology, breast cancer patients currently have a favorable prognosis, with 5-year survival rates of more than 80% for breast cancer patients under 74 years of age in most countries[ 14 ]. Treatment strategies for breast cancer are increasingly focused on individualized treatment of various complications that occur during postoperative and later radiotherapy, chemotherapy, and endocrine therapy in patients. Mugnier[ 15 ] and others have shown that antiresorptive therapy prevents aromatase inhibitor-induced bone loss in patients with early-stage breast cancer. Therefore, early identification of bone mass in breast cancer patients is essential to improve breast cancer-related bone loss. Currently, with the development of artificial intelligence, machine learning and medicine continue to be integrated. An increasing number of researchers have applied machine learning algorithms to construct models for clinical diagnosis and treatment, such as early identification of disease occurrence, determination of treatment effects, and prediction of disease prognosis. Cheng [ 16 ] et al. constructed a prediction model for microscopic breast cancer on the basis of clinical information and ultrasound parameters, and the results revealed that age, surgical margins, tumor shape, and breast density were independent factors of malignant microscopic breast cancer lesions. The AUC of this column-line graph model was 0.875. Wen [ 17 ] et al. constructed a model to predict pathological complete remission in patients with breast cancer after neoadjuvant chemotherapy, and the AUC of this model was 0.912. Our study identified 10 clinical features, including OC, radiotherapy, L1-L4 BMD,RBC, TP, 25(OH)D, Ca, memory decline, femur BMD and dizziness and blurred vision, that were significantly associated with breast cancer-related bone loss. BMD measured by dual-energy X-ray absorptiometry is the standard for assessing total bone mass, and lumbar and femoral BMD are often measured clinically to diagnose osteoporosis. Studies have shown that the incidence of pelvic instability fractures after radiotherapy for gynecologic cancers is as high as 14%, with 39.7% for sacroiliac joint fractures and 33.9% for sacral body fractures[ 18 ]. Osteocalcin is a protein derived from osteoblasts and is a marker of bone formation. Carboxy calcitonin maintains bone toughness by binding to hydroxyapatite crystals, and the serum calcitonin level increases when the BMD decreases; thus, the serum calcitonin level can be a marker for the diagnosis and screening of osteoporosis patients[ 19 ]. A study on bone loss in a Korean population revealed that red blood cell counts were reduced in men with decreased hip bone mineral density[ 20 ]. It has been shown that the serum ALB concentration has a significant effect on BMD in older women and that low serum ALB concentrations are significantly and independently associated with the incidence of osteoporosis[ 21 ]. Vitamin D deficiency may lead to high bone conversion, bone loss, and mineralization defects, and 25(OH)D is the major metabolite of vitamin D. Vitamin D deficiency is often screened for clinically by measuring serum 25(OH)D. The consensus of the Endocrine Society of the United States, Canada, and China is that blood 25(OH)D levels should be maintained at ≥ 75 nmol/L in patients with osteoporosis[ 22 ]. Researchers at the University of Hong Kong compared serum calcium levels and BMD at various sites (including L1‒L4, pelvis, femoral neck, trochanter, total hip, and whole body) in Hong Kong among southern Chinese, Mexican American, Hispanic and non-Hispanic Americans of different genetic backgrounds by Mendelian randomization analysis. In this study, the effects of smoking status, alcohol intake, serum phosphate, PTH, and 25(OH)D were excluded. The association was further strengthened, suggesting that serum calcium plays an independent role in bone metabolism[ 23 ]. Osteoporosis and Alzheimer's disease are often comorbid in the elderly population, and one study analyzed the Simple Mental State Examination (MMSE) scores and Auditory Verbal Learning Tests (AVLT tests) of the osteoporosis group and the control group and showed statistically significant results, with p values less than 0.001, and that patients with osteoporosis had poorer cognitive function. When the cognitive parameter MMSE score of 24 was used as the cutoff, the mild cognitive impairment group had lower scores on items 1–5 of the Auditory Verbal Learning Test ( p = 0.006), indicating that patients with osteoporosis are susceptible to cognitive deficits, especially declarative memory deficits[ 24 ]. Osteoporotic vertebral fractures are associated with reduced gray matter volume in specific brain regions that are responsible for memory, emotional processing, and visuospatial memory, a phenomenon that affects men more severely[ 25 ]. The results of a three-year follow-up of breast cancer patients revealed that patients with bone loss (including osteoporotic patients) accounted for 41.4% of the total at baseline and 54.5%, 60.9%, and 62.5% from the first to the third year, respectively, with an increasing trend from year to year[ 26 ]. Osteoporosis in Chinese medicine can be categorized as “bone wilting”, “bone impediment”, etc. According to the different clinical manifestations and mechanisms of disease, it can be classified into different syndrome types, among which liver-kidney depletion is most common, and the Miraculous Pivot says, “If the marrow is insufficient, the brain turns and tinnitus, and the shins are sore and dizzy.”[ 27 ] When there is insufficiency of the kidney essence, the brain has lost its nourishment, causing dizziness and blurred vision.The AdaBoost model developed in this study can be used as an auxiliary diagnostic tool for bone loss in patients with breast cancer. However, further validation in other populations is needed to determine its efficacy for practical application. Finally, we established a publicly accessible applet ( https://drive.google.com/file/d/1iCbS8SYE0HMSZsgnuYPejbJ1xHYJmskN/view?usp=drive_link ) for patients to use on their own. On this basis, we can further transform the Web tool into a physical AI robot to assist patients in medical treatment and health guidance. Although the applicability and reliability of the AdaBoost model have been verified via the ROC curve, the model still has several limitations. First, the retrospective nature of the current study involved multicenter data collection, and some patients with missing data were excluded, potentially resulting in selection bias. There is a need to scale up each region further to increase the sample size, dilute the effect of sampling error, and improve the generalization ability of the model. Second, the main features of the disease were not well characterized, as only the length of breast cancer illness, degree of lymphatic metastasis, and whether or not radiotherapy was given were considered, which inevitably led to limitations in the analysis of factors related to the disease itself. Finally, the model mainly used data from Chinese patients, and further validation of its performance and clinical utility in different ethnic groups is needed to ensure its generalizability. Conclusion In conclusion, our study developed a breast cancer bone loss model for accurate prediction of the risk of bone loss in breast cancer patients. The model demonstrated stable diagnostic performance in the training and test sets. ROC curves confirmed the feasibility of the breast cancer bone loss model in predicting bone loss. The Breast Cancer Bone Loss Risk Prediction Model is an effective tool to assist clinicians in making clinical diagnoses and treatment decisions, guiding patients to live healthy lives, and benefiting both doctors and patients in the process of managing the bone health of patients with breast cancer. Abbreviations AUC Area under curve RF Random forest LR Logistic regression KNN K nearest neighbors SVM Support Vector Machines XGB/XGBoost Extreme Gradient Boosting BT Bagging tree algorithm Ada/AdaBoost Adaptive boosting HP Hip fracture MOF Major Osteoporotic Fracture Score TBil Total bilirubin β-CTx β-Isomerized C-telopeptide LH Luteinizing hormone TP Total protein WOMAC Osteoarthritis Index Score AST Aspartate aminotransferase ALT Alanine aminotransferase OC Osteocalcin SCr Serum creatinine T Testosterone ALP Alkaline phosphatase Ca Calcium RBC Red blood cell count γ-GT γ-Glutamyl transpeptidase HGH Growth hormone 25(OH)D 25-Hydroxy vitamin D FSH Follicle-stimulating hormone PRL Prolactin BUN Blood urea nitrogen Declarations Acknowledgements The authors would like to thank the patients who participated in this study. Author contributions Meiling Chu and Guanghua Yang conceived and designed the experiments. Xiaowen Lai collected and analyzed the data, Guanghua Yang and Lifang Song evaluated and verified the data and results again, and Meiling Chu and Xiaowen Lai wrote the paper. All authors have read and agreed to the published version of the manuscript. Funding The following funds and projects supported this study: 1. Shanghai Seventh People's Hospital Medical-Industrial Cross-Innovation Special Project (QYYGJ0102); 2. The Young Medical Talents Training Program of Shanghai Pudong New Area Health Commission (PWRq2024-46). Data availability Due to patient privacy concerns within the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine Health Information Network, the primary data supporting this article cannot be shared publicly.Nevertheless, the data sets that were used in the current study are available from the corresponding author upon reasonable request. Ethical approval Ethics Approval and Consent to Participate This study was approved by the Institutional Review Board of the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine (approval number: ChiCTR2200057785). The requirement for informed consent was waived by the ethics committee due to the study’s retrospective nature. All procedures were conducted in accordance with relevant laws, regulations, and the principles outlined in the Declaration of HelsinkiConsent to participate The Institutional Review Board (IRB) of the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine waived the requirement for informed consent due to the retrospective nature of the study. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Author details 1 the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine, No. 358, Datong Road, Shanghai 200137, Shanghai, China, Phone:+86 021-58670561. † Meiling Chuand Xiaowen Lai contributed equally to this work as co-first authors. * Guanghua Yangand Lifang Song contributed equally to this work as co-corresponding authors. Co-corresponding author ： Guanghua Yang, the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine, No. 358, Datong Road, Shanghai 200137, Shanghai, China, Phone:+86 021-58670561. E-mail: [email protected] Lifang Song, the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine, No. 358, Datong Road, Shanghai 200137, Shanghai, China, Phone:+86 021-58670561. E-mail: [email protected] References BRAY F, LAVERSANNE M, SUNG H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA: a cancer journal for clinicians, 2024, 74(3): 229-263. https://doi.org/10.3322/caac.21834. XU Y, GONG M, WANG Y, et al. Global trends and forecasts of breast cancer incidence and deaths[J]. Scientific Data, 2023, 10: 334.https://doi.org/10.1038/s41597-023-02253-5. Breast Cancer Expert Committee of the National Center for Quality Control in Oncology. The standardization of the whole-cycle management of breast cancer bone health[J]. Chinese Medical Journal, 2024, 104(2): 107-124.https://doi.org/10.3760/cma.j.cn112137-20231109-01132. Breast Cancer Committee of the Chinese Anti-Cancer Association, Breast Cancer Group of the Oncology Branch of the Chinese Medical Association. Breast Cancer Diagnosis and Treatment Guidelines of the Chinese Anti-Cancer Association (2024 Edition)[J]. Chinese Journal of Cancer, 2023, 33(12): 1092-1187.https://doi.org/10.19401/j.cnki.1007-3639.2023.12.004. Li H. Clinical study on the occurrence of osteoporosis associated with breast cancer chemotherapy [D]. Chongqing Medical University, 2020[2024-12-27].https://doi.org/10.27674/d.cnki.gcyku.2020.000226. FATIMA N, LIU L, HONG S, et al. Prediction of Breast Cancer, Comparative Review of Machine Learning Techniques, and Their Analysis[J]. IEEE Access, 2020, 8: 150360-150376.https://doi.org/10.1109/ACCESS.2020.3016715. TAKEFUJI Y. Evaluating feature importance biases in logistic regression: Recommendations for robust statistical methods[J]. European Journal of Internal Medicine, 2024: S0953620524004874.https://doi.org/10.1016/j.ejim.2024.11.022. CAO T, WANG X, ZHANG H. Energy bagging tree[J]. Statistics and its interface, 2016, 9(2): 171-181.https://doi.org/10.4310/SII.2016.v9.n2.a5. RIGATTI S J. Random Forest[J]. Journal of Insurance Medicine (New York, N.Y.), 2017, 47(1): 31-39.https://doi.org/10.17849/insm-47-01-31-39.1. SARKER I H. Machine Learning: Algorithms, Real-World Applications and Research Directions[J]. Sn Computer Science, 2021, 2(3):160.https://doi.org/10.1007/s42979-021-00592-x. CHEN T, GUESTRIN C. XGBoost: A Scalable Tree Boosting System[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, NY, USA: Association for Computing Machinery, 2016: 785-794[2024-12-26].https://doi.org/10.1145/2939672.2939785. UDDIN S, HAQUE I, LU H, et al. Comparative performance analysis of K-nearest neighbour (KNN) algorithm and its different variants for disease prediction[J]. Scientific Reports, 2022, 12: 6256. https://doi.org/10.1038/s41598-022-10358-x. BATTINENI G, CHINTALAPUDI N, AMENTA F. Machine learning in medicine: Performance calculation of dementia prediction by support vector machines (SVM)[J]. Informatics in Medicine Unlocked, 2019, 16: 100200.https://doi.org/10.1016/j.imu.2019.100200. TANG D D, YE Z J, LIU W W, et al. Survival feature and trend of female breast cancer: A comprehensive review of survival analysis from cancer registration data[J]. The Breast, 2025, 79[2024-12-26].https://doi.org/10.1016/j.breast.2024.103862. MUGNIER B, GONCALVES A, DAUMAS A, et al. Prevention of aromatase inhibitor–induced bone loss with anti-resorptive therapy in post-menopausal women with early-stage breast cancer[J]. Osteoporosis International, 2023, 34(4): 703-711.https://doi.org/10.1007/s00198-023-06683-0. CHENG L L, YE F, XU T, et al. Nomogram Model for Predicting Minimal Breast Cancer Based on Clinical and Ultrasonic Characteristics[J]. International Journal of Women’s Health, 2024, 16: 2173-2184. https://doi.org/10.2147/IJWH.S482291. WEN X, CHEN J, ZHONG J, et al. Model based on ultrasound and clinicopathological characteristics for early prediction of pathological complete response to neoadjuvant chemotherapy in breast cancer[J]. Quantitative Imaging in Medicine and Surgery, 2024, 14(12): 8840-8851.https://doi.org/10.21037/qims-24-1268. SAPIENZA L G, SALCEDO M P, NING M S, et al. Pelvic Insufficiency Fractures After External Beam Radiation Therapy for Gynecologic Cancers: A Meta-analysis and Meta-regression of 3929 Patients[J]. International Journal of Radiation Oncology, Biology, Physics, 2020, 106(3): 475-484.https://doi.org/10.1016/j.ijrobp.2019.09.012. MOHAMMADI S M, SANIEE N, MOUSAVIASL S, et al. The Role of Osteocalcin in Patients with Osteoporosis: A Systematic Review[J]. Iranian Journal of Public Health, 2024, 53(11): 2432-2439.https://doi.org/10.18502/ijph.v53i11.16945. LEE J H, PARK J, KIM J H, et al. Integrative analysis of genetic and clinical risk factors for bone loss in a Korean population[J]. Bone, 2021, 147: 115910.https://doi.org/10.1016/j.bone.2021.115910. NAGAYAMA Y, EBINA K, TSUBOI H, et al. Low serum albumin concentration is associated with increased risk of osteoporosis in postmenopausal patients with rheumatoid arthritis[J]. Journal of Orthopedic Science, 2022, 27(6): 1283-1290. https://doi.org/10.1016/j.jos.2021.08.018. LEI S, ZHANG X, SONG L, et al. Expert consensus on vitamin D in osteoporosis[J]. Annals of Joint, 2025, 10: 1.https://doi.org/10.21037/aoj-24-48. LI G H Y, ROBINSON-COHEN C, SAHNI S, et al. Association of Genetic Variants Related to Serum Calcium Levels with Reduced Bone Mineral Density[J]. The Journal of Clinical Endocrinology and Metabolism, 2019, 105(3): e328-e336. https://doi.org/10.1210/clinem/dgz088. PENG W, LI Z, GUAN Y, et al. A study of cognitive functions in female elderly patients with osteoporosis: a multicenter cross-sectional study[J]. Aging & Mental Health, 2016, 20(6): 647-654.https://doi.org/10.1080/13607863.2015.1033680. NAKAJIMA K, HORII C, KODAMA H, et al. Association between vertebral fractures and brain volume: insights from a community cohort study[J]. Osteoporosis International, 2025[2025-03-24].https://doi.org/10.1007/s00198-025-07403-6. KIM S H, CHO Y U, KIM S J, et al. Changes in Bone Mineral Density in Women With Breast Cancer: A Prospective Cohort Study[J]. Cancer Nursing, 2019, 42(2): 164-172.https://doi.org/10.1097/NCC.0000000000000586. Wang ZG. Ling Shu Jing [M]. Beijing: China Press of Traditional Chinese Medicine, 2022: 156. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 26 May, 2025 Reviews received at journal 25 May, 2025 Reviewers agreed at journal 23 May, 2025 Reviewers agreed at journal 23 May, 2025 Reviews received at journal 22 May, 2025 Reviewers agreed at journal 21 May, 2025 Reviews received at journal 21 May, 2025 Reviewers agreed at journal 20 May, 2025 Reviewers agreed at journal 06 May, 2025 Reviewers agreed at journal 05 May, 2025 Reviewers agreed at journal 03 May, 2025 Reviewers agreed at journal 02 May, 2025 Reviewers invited by journal 30 Apr, 2025 Editor invited by journal 22 Apr, 2025 Editor assigned by journal 21 Apr, 2025 Submission checks completed at journal 21 Apr, 2025 First submitted to journal 15 Apr, 2025 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-6455887","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":451388813,"identity":"b49bd539-bf6b-400a-b4a9-e4340c4896df","order_by":0,"name":"Meiling Chu","email":"","orcid":"","institution":"the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Meiling","middleName":"","lastName":"Chu","suffix":""},{"id":451388814,"identity":"b695c379-aea9-40ec-8953-dcaeade4dd72","order_by":1,"name":"Xiaowen Lai","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA70lEQVRIiWNgGAWjYBACPgbGBoYEMJOx4UACjw0PP38Dfi1sCC3MjQceyKTJSM44QEgLHLA3H3xgc9jGoCGBgBb2w20SD2psEvulG4EOyznPY8BwgPHDxxw8WngSmw0SjqUlzpxzEKjlzG0ec+YGZsmZ2/A5LLHxQQLb4dwNNxIbDiT23OaxbDjAxsyLTwv/Q6Dh/w7n7gdr+XeOx+BAAgEtEkBbEtuAtkgkggL5ADFaHjYbJPal1c+4AdaSzCM542AzXr/w86c/k/zxzcaYf0b6448/eOzs+fmbD374iEcLNgCM3FEwCkbBKBgFlAEAnKhag8E0KDYAAAAASUVORK5CYII=","orcid":"","institution":"the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine","correspondingAuthor":true,"prefix":"","firstName":"Xiaowen","middleName":"","lastName":"Lai","suffix":""},{"id":451388815,"identity":"949135eb-e977-43f6-bae3-5a9fbee71bd7","order_by":2,"name":"Guanghua Yang","email":"","orcid":"","institution":"the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Guanghua","middleName":"","lastName":"Yang","suffix":""},{"id":451388816,"identity":"b3bb318f-e4f0-4e77-87a6-07d3456c3513","order_by":3,"name":"Lifang Song","email":"","orcid":"","institution":"the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Lifang","middleName":"","lastName":"Song","suffix":""}],"badges":[],"createdAt":"2025-04-15 14:38:23","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6455887/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6455887/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":82276290,"identity":"9dc2a5f2-6005-4225-8cca-282b63adf456","added_by":"auto","created_at":"2025-05-08 14:45:54","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":70966,"visible":true,"origin":"","legend":"\u003cp\u003eSHapley additive interpretation plot: effects of clinical features in random forest, logistic regression, support vector machine, eXtreme gradient boosting, the Bagging Tree algorithm, k nearest neighbors, and adaptive boosting for identifying breast cancer-related bone loss. A. RF, B. LR, C. SVM, D. XGB, E. Bagging Tree, F. KNN, G. AdaBoost. L1--L4 BMD: Bone density T value of the first lumbar vertebra to the fourth lumbar vertebra; Femur BMD: Bone density T value of the femur; HP: Hip fracture; MOF: Major Osteoporotic Fracture Score, i.e., probability of occurrence of a major osteoporotic fracture; TBil: Total bilirubin; β-CTx: β-Isomerized C-telopeptide; LH: Luteinizing hormone; TP: Total protein; WOMAC: Osteoarthritis Index Score; AST: Aspartate aminotransferase; ALT: Alanine aminotransferase; OC: Osteocalcin; SCr: Serum creatinine; T: Testosterone; ALP: Alkaline phosphatase; Ca: Calcium; RBC: Red blood cell count; γ-GT: γ-Glutamyl transpeptidase; HGH: Growth hormone; 25(OH)D: 25-Hydroxy vitamin D; FSH: Follicle-stimulating hormone; PRL: Prolactin; BUN: Blood urea nitrogen.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6455887/v1/e724572d2dadc0d83b2810f6.png"},{"id":82278085,"identity":"e726c229-79f2-41f5-9082-ef78c0e924bd","added_by":"auto","created_at":"2025-05-08 14:53:54","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":247859,"visible":true,"origin":"","legend":"\u003cp\u003e(A) ROC curves of seven machine learning models.(B) Diagnostic performance of seven machine learning models in the training cohort and test cohort for predicting bone loss associated with breast cancer. Logi: logistic regression, BT: bagging tree algorithm, RF: random forest, Ada: AdaBoost, XGB: eXtreme gradient boosting, KNN: k nearest neighbor, SVM: support vector machine.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6455887/v1/7ee869dc0f5be452da200fe7.png"},{"id":82279338,"identity":"1166fce5-4794-4995-a019-bc1d418b00fc","added_by":"auto","created_at":"2025-05-08 15:01:55","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1227610,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6455887/v1/eae9d941-e830-4687-8081-9342583b5525.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A study on the development of a machine learning-based clinical prediction model for bone health management in breast cancer patients","fulltext":[{"header":"Introduction","content":"\u003cp\u003eAccording to the Global Cancer Statistics 2022 published by the International Agency for Research on Cancer (IARC) of the World Health Organization, breast cancer has become the malignant tumor with the highest incidence in women worldwide[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Studies have shown that the global incidence of breast cancer will increase every year from 2020 to 2050, and 4,781,849 cases of breast cancer and 1,503,694 deaths from breast cancer are expected to occur worldwide in 2050[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Notably, bone loss caused by age, disease, treatment, hormone level changes, and other factors during the diagnosis and treatment of breast cancer patients has become a concomitant symptom of the high incidence of breast cancer[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e], which affects the quality of life of patients and their recovery from the disease.In 2023, the China Oncology Healthcare Management Conference promulgated the inaugural set of guidelines for standardized management of the entire cycle of bone health in breast cancer. Bone safety management for breast cancer patients refers to the early assessment, prevention, diagnosis, treatment, and rehabilitation management of bone loss, osteoporosis, and fracture risk events during disease treatment and follow-up[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Early bone health management for patients with breast cancer includes the prevention and treatment of bone loss caused by tumors and the prevention of bone metastasis. The Chinese Anti-Cancer Society Breast Cancer Diagnostic and Treatment Guidelines and Specifications (Version 2024) suggest making staging diagnoses and corresponding treatment plans for breast cancer staging criteria on the basis of clinical examination and surgical pathology results[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. The musculoskeletal-related side effects of endocrine therapy, radiotherapy, chemotherapy, and other posttreatments for breast cancer seriously affect patients' quality of life and medication compliance. The incidence of abnormal bone mass in breast cancer patients in China reaches 77.7%, and the incidence of osteoporosis is 30.5%[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Research has shown that the incidence of bone metastasis in advanced stages of breast cancer is 65%~75%[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. The essence of Chinese medicine lies in holistic concepts and evidence-based treatment. Chinese medicine emphasizes that the human body is an organic whole and that all parts are interrelated and affect each other. Through inspection, auscultation-olfaction, inquiry, and palpation, the four diagnostic methods are combined to analyze the cause and mechanism of the disease according to the patient's constitution to formulate a personalized treatment plan, which focuses not only on the alleviation of symptoms but also on eliminating the causes of the disease and conditioning the constitution. Therefore, a medical support system that combines Chinese and Western medicine is increasingly favored for the local recurrence and metastasis of breast cancer and its treatment. Currently, Fatima et al.[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e] have conducted a comprehensive comparative analysis of machine learning techniques for breast cancer prediction, offering detailed insights into the algorithms employed by various machine learning approaches in this domain. However, developing efficient methods for breast cancer-related bone loss prediction with both iterative learning capabilities and interpretability to continuously improve model efficacy is challenging. We aim to design and develop an applet for clinical practice that facilitates an intelligent service system for health counseling, intelligent diagnosis, assisted diagnosis, and treatment decision-making for the bone health management of patients with breast cancer. To play an active role in the management of bone safety in the whole management and all-round management chain of breast cancer and to be widely used in clinical practice and welfare society.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy population\u003c/h2\u003e \u003cp\u003eThis multicenter retrospective study included breast cancer patients from March 2022 to December 2024 from three tertiary-level hospitals in China: the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine, the Longhua Hospital affiliated with Shanghai University of Traditional Chinese Medicine, and the Shengjing Hospital Affiliated with China Medical University.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eInclusion and Exclusion criteria\u003c/h3\u003e\n\u003cp\u003eThe inclusion criteria were as follows: (1) aged between 18 and 80 years; (2) KPS score\u0026thinsp;\u0026ge;\u0026thinsp;60; and (3) diagnosed with breast cancer by pathology. The exclusion criteria were as follows: (1) bone mineral diseases, such as deformational osteitis, osteogenesis imperfecta, osteochondrosis, etc.; (2) diseases affecting bone metabolism, such as Cushing's syndrome, rheumatoid or rheumatoid arthritis, etc.; (3) hormone replacement therapy used in the past 6 months, such as glucocorticoids, parathyroid hormone, etc.; (4) surgical treatment for limb joints carried out in the past 6 months; (5) severe cardiovascular, liver, renal or hematopoietic systems and other serious primary diseases; (6) psychiatric patients with cognitive loss or mental abnormality; and (7) insufficient clinical information in the medical file.\u003c/p\u003e\n\u003ch3\u003eData collection\u003c/h3\u003e\n\u003cp\u003eBefore initiating data collection, all relevant staff involved were trained on the data extraction form. The medical records of eligible patients were meticulously reviewed. Clinical and laboratory data (including demographic characteristics, TCM symptom scales, routine blood tests, paraffin pathology reports of breast cancer, and T values of bone density measured by dual-energy X-ray absorptiometry) were systematically extracted into standardized forms for this project. To ensure the relevance and consistency of the data, 30% of the randomly extracted data were evaluated and validated by the principal. Units of measurement were standardized across all laboratories, taking into account differences in units across participating centers. Extreme outliers, either significantly large or significantly small values, were flagged for review. These outliers were reevaluated by the program principal or designated attending physician to determine the validity of the data and to rule out data entry errors.\u003c/p\u003e\n\u003ch3\u003eOverview of machine learning models\u003c/h3\u003e\n\u003cp\u003eWe developed and compared seven models, including LR, the Bagging Tree algorithm, RF, AdaBoost, XGBoost, KNN, and SVM, for predicting the risk of bone loss and bone metastasis in patients with breast cancer. Logistic regression is a method widely used for binary outcomes, especially in the social sciences, epidemiology, and machine learning[\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. The Bagging Tree algorithm belongs to a wider range of integrated methods that improve the stability and accuracy of tree-based methods used in nonparametric regression, mainly through voting or model averaging[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Random forest is a further improvement in classification and regression tree models that integrates multiple decision trees to better identify meaningful interactions and nonlinear effects in predictor variables[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. AdaBoost is an integrated learning process that employs an iterative approach to improve poor-quality classifiers by learning from errors[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. XGBoost is a scalable machine for tree boosting learning systems for dealing with sparse data, which increases the speed of learning and thus speeds up model exploration [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. KNN is a supervised machine learning algorithm applied for classification purposes that can be trained to predict the classification of unknown data by training the features and labels of the known data, and it is now widely used for disease prediction[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. SVM is a discriminative classifier characterized by an optimal hyperplane to maximize the interval between different classes[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e\n\u003ch3\u003ePrimary objective\u003c/h3\u003e\n\u003cp\u003eThe primary objective of this study was to construct a machine learning model for rapid adjuvant diagnosis and treatment for the bone health management of breast cancer patients.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003ePatient data are presented as categorical or continuous variables. The Kolmogorov‒Smirnov test was used to assess whether the data were normally distributed. For normally distributed continuous variables, the data are presented as the means\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviations and were compared via t tests. If continuous variables did not follow a normal distribution, the Mann‒Whitney U test was used, and the results are expressed as the median (interquartile spacing). Categorical data are expressed as numbers and frequencies and were compared via the chi-square test or Fisher's exact test. All the statistical tests were two-sided, and p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 indicated statistical significance. SPSS software version 25.0 was used in this study.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cp\u003eThis study included a total of 66 candidate predictors, including basic patient information, disease diagnosis and treatment,the Fracture Risk Assessment Tool (FRAX) and the Western Ontario McMaster University Osteoarthritis Index Scale (WOMA), involving 25 characteristic variables, as well as 25 blood index testing items and 16 TCM symptom elements.The 66 variables from the training cohort were incorporated into 7 machine algorithms, namely, the LR, Bagging Tree, RF, AdaBoost, XGBoost, KNN, and SVM algorithms, to generate a prediction model for predicting bone loss in patients with breast cancer, and the prediction model was used to provide a quantitative description of the overall relationship between breast cancer-related bone loss and all 66 features. The features were characterized using the Shapley significance analysis, and 10 highly sensitive features were screened out. In this process, the \"contribution\" of each feature to the model is calculated on the basis of the marginal contribution of each feature and the number of occurrences of the feature, and the optimal algorithm is selected after the performance differences between different machine learning algorithms are compared. The performance of the model is evaluated usingvia two methods: the ROC curve and the confusion matrix. The confusion matrix shows the number of correctly predicted samples and the number of incorrectly predicted samples in each category in the form of a matrix (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\u003eConfusion matrix\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCondition positive\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCondition positive\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTest outcome positive\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTrue Positive (TP)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFalse Positive (FP)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTest outcome negative\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eFalse Negative (FN)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eTrue Negative (TN)\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\u003eAccuracy: represents the proportion of the correct sample size to the total sample size.\u003c/p\u003e \u003cp\u003e\u003cimg src=\"data:image/png;base64,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\" width=\"235\" height=\"57\"\u003e\u003c/p\u003e\n\u003ch3\u003eRole of the funding source\u003c/h3\u003e\n\u003cp\u003eThe funder of this study had no role in the research design, data collection, data analysis, data interpretation, or report writing.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e1 Patient characteristics\u003c/h2\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\u003eBaseline characteristics of the cohort\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=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" 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\u0026nbsp;\u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNormal group (n\u0026thinsp;=\u0026thinsp;61)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOsteopenia group (n\u0026thinsp;=\u0026thinsp;78)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e Value\u003c/p\u003e \u003cp\u003e(2-sided)\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=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e47.49\u0026thinsp;\u0026plusmn;\u0026thinsp;8.57\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e49.87\u0026thinsp;\u0026plusmn;\u0026thinsp;9.038\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.118\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBMI (kg/m\u003csup\u003e2)\u003c/sup\u003e\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e23.94\u0026thinsp;\u0026plusmn;\u0026thinsp;3.69\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e23.14\u0026thinsp;\u0026plusmn;\u0026thinsp;3.128\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.157\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eEducation\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHigh school and below\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e32(52.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41(52.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.99\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMore than university\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e29(47.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e37(47.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge at menarche\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13.77\u0026thinsp;\u0026plusmn;\u0026thinsp;1.383\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e14.09\u0026thinsp;\u0026plusmn;\u0026thinsp;1.321\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.106\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eMenopause\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e39(63.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e41(52.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e0.178\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eYes\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e22(36.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e37(47.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eNumber of pregnancies\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6(9.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1(1.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e0.072\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOnce or twice\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e40(65.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e57(73.1%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMore than twice\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e15(24.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e20(25.6%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003eNumber of production times\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNo\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6(9.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3(3.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\" morerows=\"2\" rowspan=\"3\"\u003e \u003cp\u003e0.339\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOnce\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e41(67.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e58(74.4%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTwice and more\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e14(23%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e17(21.8%)\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDuration of breast cancer (months)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e17.11\u0026thinsp;\u0026plusmn;\u0026thinsp;10.767\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e26.54\u0026thinsp;\u0026plusmn;\u0026thinsp;15.965\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFamily history of cancer\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e18(29.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e24(30.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.872\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLymph node metastasis\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e27(44.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e32(41.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.702\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTargeted therapy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e9(14.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e17(21.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.291\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRadiotherapy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e36(59.0%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e47(60.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.882\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eChemotherapy\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e27(44.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e39(50%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.501\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDrink coffee\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e13(21.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e13(16.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.486\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eExercise\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e25(41%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e36(46.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.542\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHistory of fractures\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e6(9.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e10(12.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.584\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHistory of parental hip fracture\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e7(11.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e17(21.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.11\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eL1-L4 BMD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.177\u0026thinsp;\u0026plusmn;\u0026thinsp;0.910\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-1.468\u0026thinsp;\u0026plusmn;\u0026thinsp;0.842\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFemur BMD\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e-0.151\u0026thinsp;\u0026plusmn;\u0026thinsp;0.569\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-1.124\u0026thinsp;\u0026plusmn;\u0026thinsp;0.828\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eFRAX\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMajor Osteoporotic Fracture\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e2.303\u0026thinsp;\u0026plusmn;\u0026thinsp;1.741\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e3.705\u0026thinsp;\u0026plusmn;\u0026thinsp;3.255\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eHip Fracture\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e0.336\u0026thinsp;\u0026plusmn;\u0026thinsp;1.064\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.851\u0026thinsp;\u0026plusmn;\u0026thinsp;1.63\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e\u0026lt;0.001\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWOMAC\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e26.21\u0026thinsp;\u0026plusmn;\u0026thinsp;15.80\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e30.13\u0026thinsp;\u0026plusmn;\u0026thinsp;15.485\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e \u003cp\u003e0.3\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\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\u003eAnalysis of blood indicators in the study population\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=\"\u0026plusmn;\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\"\u0026plusmn;\" 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\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNormal group (n\u0026thinsp;=\u0026thinsp;61)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOsteopenia group (n\u0026thinsp;=\u0026thinsp;78)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e Value (2-sided)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGrowth hormone (ug/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.879\u0026thinsp;\u0026plusmn;\u0026thinsp;1.125\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e1.015\u0026thinsp;\u0026plusmn;\u0026thinsp;1.329\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.551\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEstradiol (pg/mL)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e23.572\u0026thinsp;\u0026plusmn;\u0026thinsp;97.177\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e20.913\u0026thinsp;\u0026plusmn;\u0026thinsp;85.362\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.273\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFollicle-stimulating hormone (IU/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e32.404\u0026thinsp;\u0026plusmn;\u0026thinsp;33.001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e34.472\u0026thinsp;\u0026plusmn;\u0026thinsp;31.274\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.257\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLuteinizing hormone (IU/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e11.622\u0026thinsp;\u0026plusmn;\u0026thinsp;14.783\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e12.56\u0026thinsp;\u0026plusmn;\u0026thinsp;15.194\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.686\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProlactin (ng/mL)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e190.966\u0026thinsp;\u0026plusmn;\u0026thinsp;138.584\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e149.568\u0026thinsp;\u0026plusmn;\u0026thinsp;96.881\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.135\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eProgesterone (ng/mL)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e0.625\u0026thinsp;\u0026plusmn;\u0026thinsp;0.349\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e0.781\u0026thinsp;\u0026plusmn;\u0026thinsp;0.555\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.275\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTestosterone (nmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e1.142\u0026thinsp;\u0026plusmn;\u0026thinsp;1.233\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e1.245\u0026thinsp;\u0026plusmn;\u0026thinsp;2.644\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.569\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e25-Hydroxyvitamin D (nmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e64.83\u0026thinsp;\u0026plusmn;\u0026thinsp;24.39\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e71.388\u0026thinsp;\u0026plusmn;\u0026thinsp;22.233\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.10\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eOsteocalcin (ng/mL)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e26.039\u0026thinsp;\u0026plusmn;\u0026thinsp;10.665\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e23.641\u0026thinsp;\u0026plusmn;\u0026thinsp;9.483\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.14\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eβ-Isomerized C-telopeptide (pg/mL)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e639.700\u0026thinsp;\u0026plusmn;\u0026thinsp;250.535\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e566.352\u0026thinsp;\u0026plusmn;\u0026thinsp;273.333\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.106\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlanine aminotransferase (U/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e21.802\u0026thinsp;\u0026plusmn;\u0026thinsp;15.195\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e21.923\u0026thinsp;\u0026plusmn;\u0026thinsp;14.979\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.963\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAspartate aminotransferase (U/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e22.367\u0026thinsp;\u0026plusmn;\u0026thinsp;11.842\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e21.617\u0026thinsp;\u0026plusmn;\u0026thinsp;7.124\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.553\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSerum creatinine (\u0026micro;mol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e52.362\u0026thinsp;\u0026plusmn;\u0026thinsp;13.343\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e54.466\u0026thinsp;\u0026plusmn;\u0026thinsp;10.673\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.32\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eBlood urea nitrogen (mmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e5.948\u0026thinsp;\u0026plusmn;\u0026thinsp;8.560\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e4.984\u0026thinsp;\u0026plusmn;\u0026thinsp;1.184\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.609\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eγ-Glutamyl transpeptidase (U/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e28.104\u0026thinsp;\u0026plusmn;\u0026thinsp;19.926\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e25.715\u0026thinsp;\u0026plusmn;\u0026thinsp;17.79\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.724\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAlkaline phosphatase (U/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e79.590\u0026thinsp;\u0026plusmn;\u0026thinsp;26.444\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e87.997\u0026thinsp;\u0026plusmn;\u0026thinsp;35.428\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.198\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal bilirubin (\u0026micro;mol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e12.421\u0026thinsp;\u0026plusmn;\u0026thinsp;5.688\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e13.41\u0026thinsp;\u0026plusmn;\u0026thinsp;7.188\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.158\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCalcium (mmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e2.389\u0026thinsp;\u0026plusmn;\u0026thinsp;0.128\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e2.399\u0026thinsp;\u0026plusmn;\u0026thinsp;0.136\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.587\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal protein (g/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e72.038\u0026thinsp;\u0026plusmn;\u0026thinsp;6.523\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e73.677\u0026thinsp;\u0026plusmn;\u0026thinsp;3.955\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.392\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eWhite blood cell count (\u0026times;10\u003csup\u003e9\u003c/sup\u003e/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e5.485\u0026thinsp;\u0026plusmn;\u0026thinsp;1.738\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e4.909\u0026thinsp;\u0026plusmn;\u0026thinsp;1.03\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.049\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eNeutrophil count (\u0026times;10\u003csup\u003e9\u003c/sup\u003e/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e3.318\u0026thinsp;\u0026plusmn;\u0026thinsp;1.572\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e2.794\u0026thinsp;\u0026plusmn;\u0026thinsp;0.784\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.06\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLymphocyte count (\u0026times;10\u003csup\u003e9\u003c/sup\u003e/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e1.775\u0026thinsp;\u0026plusmn;\u0026thinsp;0.556\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e2.094\u0026thinsp;\u0026plusmn;\u0026thinsp;3.317\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.914\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eRed blood cell count (\u0026times;10\u003csup\u003e12\u003c/sup\u003e/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e4.399\u0026thinsp;\u0026plusmn;\u0026thinsp;0.443\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e4.477\u0026thinsp;\u0026plusmn;\u0026thinsp;0.339\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.502\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHemoglobin (g/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e132.541\u0026thinsp;\u0026plusmn;\u0026thinsp;13.581\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e134.718\u0026thinsp;\u0026plusmn;\u0026thinsp;8.197\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.74\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePlatelet count (\u0026times;10\u003csup\u003e9\u003c/sup\u003e/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c2\"\u003e \u003cp\u003e213.72\u0026thinsp;\u0026plusmn;\u0026thinsp;54.667\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\"\u0026plusmn;\" colname=\"c3\"\u003e \u003cp\u003e205.33\u0026thinsp;\u0026plusmn;\u0026thinsp;43.188\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.314\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\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\u003eAnalysis of the Chinese medicine symptoms of the study subjects\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=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" 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\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNormal group (n\u0026thinsp;=\u0026thinsp;61)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOsteopenia group (n\u0026thinsp;=\u0026thinsp;78)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u003cem\u003ep\u003c/em\u003e Value (2-sided)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003ePain\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e56(91.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e75(96.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.299\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLumbar-knee soreness with weakness\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e53(86.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e65(83.3%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.562\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGait disturbance\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e39(63.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e53(67.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.62\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCramping in the lower extremities\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e19(31.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e34(43.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.134\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMorning stiffness; numbness in the finger joints\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e48(78.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e71(91%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.040\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCold intolerance and chilly extremities\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e34(55.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e51(65.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.247\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eIncreased nocturia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e46(75.4%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e56(71.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.632\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLoose stool\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e26(42.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e42(53.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.189\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eShortness of breath and fatigue\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e47(77%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e70(89.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.042\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHot flashes and sweating\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e51(83.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e68(87.2%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.551\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTidal fever and night sweats\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e45(73.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e57(73.1%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.927\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDry throat and mouth\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e47(77%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e60(76.9%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.986\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDizziness and blurred vision\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e40(65.6%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e55(70.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.534\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTinnitus\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31(50.8%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e39(50%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.924\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMemory decline\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e54(88.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e77(98.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eInsomnia\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e48(78.7%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e62(79.5%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e \u003cp\u003e0.908\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\u003eA total of 139 study participants were categorized into 61 individuals in the normal bone mass group (T\u0026thinsp;\u0026gt;\u0026thinsp;1.0) and 78 individuals in the reduced bone mass group (including those with osteoporosis) (T\u0026thinsp;\u0026le;\u0026thinsp;1.0) on the basis of the T value of BMD. The baseline characteristics of the study sample are shown in Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e. The duration of breast cancer was longer in the reduced bone mass group than in the normal bone mass group, with means of 17.11\u0026thinsp;\u0026plusmn;\u0026thinsp;10.767 and 26.54\u0026thinsp;\u0026plusmn;\u0026thinsp;15.965, respectively. The overall probability of osteoporosis and the probability of hip fracture were statistically significant in both groups, as calculated by the Fracture Risk Assessment Tool (both \u003cem\u003ep\u003c/em\u003e values were less than 0.001). The white blood cell count was significantly different (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.049) between the two groups (Table\u0026nbsp;\u003cspan refid=\"Tab3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). A comparison of the TCM evidence scale between the two groups also revealed favorable results, with the incidence of morning stiffness, numbness of the knuckles, shortness of breath, and memory loss in the normal bone mass group (78.7%, 77%, and 88.5%, respectively) and the reduced bone mass group (91%, 89.7%, and 98.7%, respectively) (Table\u0026nbsp;\u003cspan refid=\"Tab4\" class=\"InternalRef\"\u003e4\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003e2 Performance of each predictive model in identifying breast cancer-associated bone loss\u003c/h2\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eSixty-six variables from the training cohort were integrated into the ML algorithm to create predictive models, and 10 variables with high significance were evaluated. The final RF model included L1‒L4 BMD, femur BMD, duration of breast cancer, MOF, HF, TBil, β-CTx, LH, TP, and WOMAC scores (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA), which had poorer accuracy than the other models in the test set of 0.810 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). The variables included in the LR model included L1-L4 BMD, femur BMD, dizziness and blurred vision, shortness of breath and fatigue, AST, OC, insomnia, increased nocturia, SCr, and testosterone (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB), and the model's accuracies in the training and test sets were 1.00 and 0.857, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). The SVM model included L1\u0026ndash;L4 BMD, insomnia, dizziness and blurred vision, the WOMAC score, femur BMD, AST, OC, shortness of breath and fatigue, T, ALP (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC), and the model had an accuracy of 0.833 in the test set (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). The XGBoost model included L1\u0026ndash;L4 BMD, femur BMD, Ca, OC, radiotherapy, memory decline, ALT, education, RBC, and TBil (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eD), which performed well on both the training and test sets, with accuracies of 1.00 and 0.929, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). The Bagging Tree model included L1‒L4 BMD, femur BMD, SCr, Ca, WOMAC score, γ-GT, HGH, ALT, β-CTx, and OC (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eE). The accuracy of the bagging tree model accuracy was 0.952 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). KNN included L1-L4 BMD, femur BMD, FSH, shortness of breath and fatigue, PRL, BUN, education, age at menarche, duration of breast cancer, and OC (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eF), and the training set accuracy of this model showed poor performance at 0.887 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). The AdaBoost model included OC, radiotherapy, L1-L4 BMD,RBC, TP, 25(OH)D, Ca, memory decline, femur BMD and dizziness and blurred vision (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eG), and the AdaBoost model had the best performance in the test set, with an accuracy of 0.952 (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB). The AUC of the RF, LR, Bagging Tree, AdaBoost, XGBoost, KNN, and SVM models were 0.957, 0.907, 0.963, 0.995, 0.991, 0.860, and 0.868, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA), and the best AUC was obtained by AdaBoost.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this study, we developed and validated a model based on a machine learning algorithm to predict breast cancer-related bone loss. The AdaBoost model was selected as the final model, which had an AUC of 0.995 and an accuracy of 0.952. These findings demonstrated the accuracy and stability of the model in predicting bone loss in patients with breast cancer. Our study suggests that the AdaBoost model has the potential to identify the degree of risk of developing osteoporosis in breast cancer patients, which can help clinicians grasp the risk of bone loss occurring during the management of breast cancer for further examination and assist in treatment decisions.\u003c/p\u003e \u003cp\u003eOwing to the emphasis on breast cancer screening and advances in medical technology, breast cancer patients currently have a favorable prognosis, with 5-year survival rates of more than 80% for breast cancer patients under 74 years of age in most countries[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Treatment strategies for breast cancer are increasingly focused on individualized treatment of various complications that occur during postoperative and later radiotherapy, chemotherapy, and endocrine therapy in patients. Mugnier[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e] and others have shown that antiresorptive therapy prevents aromatase inhibitor-induced bone loss in patients with early-stage breast cancer. Therefore, early identification of bone mass in breast cancer patients is essential to improve breast cancer-related bone loss.\u003c/p\u003e \u003cp\u003eCurrently, with the development of artificial intelligence, machine learning and medicine continue to be integrated. An increasing number of researchers have applied machine learning algorithms to construct models for clinical diagnosis and treatment, such as early identification of disease occurrence, determination of treatment effects, and prediction of disease prognosis. Cheng [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e] et al. constructed a prediction model for microscopic breast cancer on the basis of clinical information and ultrasound parameters, and the results revealed that age, surgical margins, tumor shape, and breast density were independent factors of malignant microscopic breast cancer lesions. The AUC of this column-line graph model was 0.875. Wen [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e] et al. constructed a model to predict pathological complete remission in patients with breast cancer after neoadjuvant chemotherapy, and the AUC of this model was 0.912.\u003c/p\u003e \u003cp\u003eOur study identified 10 clinical features, including OC, radiotherapy, L1-L4 BMD,RBC, TP, 25(OH)D, Ca, memory decline, femur BMD and dizziness and blurred vision, that were significantly associated with breast cancer-related bone loss. BMD measured by dual-energy X-ray absorptiometry is the standard for assessing total bone mass, and lumbar and femoral BMD are often measured clinically to diagnose osteoporosis. Studies have shown that the incidence of pelvic instability fractures after radiotherapy for gynecologic cancers is as high as 14%, with 39.7% for sacroiliac joint fractures and 33.9% for sacral body fractures[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Osteocalcin is a protein derived from osteoblasts and is a marker of bone formation. Carboxy calcitonin maintains bone toughness by binding to hydroxyapatite crystals, and the serum calcitonin level increases when the BMD decreases; thus, the serum calcitonin level can be a marker for the diagnosis and screening of osteoporosis patients[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. A study on bone loss in a Korean population revealed that red blood cell counts were reduced in men with decreased hip bone mineral density[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. It has been shown that the serum ALB concentration has a significant effect on BMD in older women and that low serum ALB concentrations are significantly and independently associated with the incidence of osteoporosis[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Vitamin D deficiency may lead to high bone conversion, bone loss, and mineralization defects, and 25(OH)D is the major metabolite of vitamin D. Vitamin D deficiency is often screened for clinically by measuring serum 25(OH)D. The consensus of the Endocrine Society of the United States, Canada, and China is that blood 25(OH)D levels should be maintained at \u0026ge;\u0026thinsp;75 nmol/L in patients with osteoporosis[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Researchers at the University of Hong Kong compared serum calcium levels and BMD at various sites (including L1‒L4, pelvis, femoral neck, trochanter, total hip, and whole body) in Hong Kong among southern Chinese, Mexican American, Hispanic and non-Hispanic Americans of different genetic backgrounds by Mendelian randomization analysis. In this study, the effects of smoking status, alcohol intake, serum phosphate, PTH, and 25(OH)D were excluded. The association was further strengthened, suggesting that serum calcium plays an independent role in bone metabolism[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Osteoporosis and Alzheimer's disease are often comorbid in the elderly population, and one study analyzed the Simple Mental State Examination (MMSE) scores and Auditory Verbal Learning Tests (AVLT tests) of the osteoporosis group and the control group and showed statistically significant results, with \u003cem\u003ep\u003c/em\u003e values less than 0.001, and that patients with osteoporosis had poorer cognitive function. When the cognitive parameter MMSE score of 24 was used as the cutoff, the mild cognitive impairment group had lower scores on items 1\u0026ndash;5 of the Auditory Verbal Learning Test (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.006), indicating that patients with osteoporosis are susceptible to cognitive deficits, especially declarative memory deficits[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. Osteoporotic vertebral fractures are associated with reduced gray matter volume in specific brain regions that are responsible for memory, emotional processing, and visuospatial memory, a phenomenon that affects men more severely[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. The results of a three-year follow-up of breast cancer patients revealed that patients with bone loss (including osteoporotic patients) accounted for 41.4% of the total at baseline and 54.5%, 60.9%, and 62.5% from the first to the third year, respectively, with an increasing trend from year to year[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Osteoporosis in Chinese medicine can be categorized as \u0026ldquo;bone wilting\u0026rdquo;, \u0026ldquo;bone impediment\u0026rdquo;, etc. According to the different clinical manifestations and mechanisms of disease, it can be classified into different syndrome types, among which liver-kidney depletion is most common, and \u003cem\u003ethe Miraculous Pivot\u003c/em\u003e says, \u0026ldquo;If the marrow is insufficient, the brain turns and tinnitus, and the shins are sore and dizzy.\u0026rdquo;[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] When there is insufficiency of the kidney essence, the brain has lost its nourishment, causing dizziness and blurred vision.The AdaBoost model developed in this study can be used as an auxiliary diagnostic tool for bone loss in patients with breast cancer. However, further validation in other populations is needed to determine its efficacy for practical application. Finally, we established a publicly accessible applet (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://drive.google.com/file/d/1iCbS8SYE0HMSZsgnuYPejbJ1xHYJmskN/view?usp=drive_link\u003c/span\u003e\u003cspan address=\"https://drive.google.com/file/d/1iCbS8SYE0HMSZsgnuYPejbJ1xHYJmskN/view?usp=drive_link\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) for patients to use on their own. On this basis, we can further transform the Web tool into a physical AI robot to assist patients in medical treatment and health guidance.\u003c/p\u003e \u003cp\u003eAlthough the applicability and reliability of the AdaBoost model have been verified via the ROC curve, the model still has several limitations. First, the retrospective nature of the current study involved multicenter data collection, and some patients with missing data were excluded, potentially resulting in selection bias. There is a need to scale up each region further to increase the sample size, dilute the effect of sampling error, and improve the generalization ability of the model. Second, the main features of the disease were not well characterized, as only the length of breast cancer illness, degree of lymphatic metastasis, and whether or not radiotherapy was given were considered, which inevitably led to limitations in the analysis of factors related to the disease itself. Finally, the model mainly used data from Chinese patients, and further validation of its performance and clinical utility in different ethnic groups is needed to ensure its generalizability.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn conclusion, our study developed a breast cancer bone loss model for accurate prediction of the risk of bone loss in breast cancer patients. The model demonstrated stable diagnostic performance in the training and test sets. ROC curves confirmed the feasibility of the breast cancer bone loss model in predicting bone loss. The Breast Cancer Bone Loss Risk Prediction Model is an effective tool to assist clinicians in making clinical diagnoses and treatment decisions, guiding patients to live healthy lives, and benefiting both doctors and patients in the process of managing the bone health of patients with breast cancer.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cp\u003eAUC\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Area under curve\u003c/p\u003e\n\u003cp\u003eRF \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Random forest\u003c/p\u003e\n\u003cp\u003eLR\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Logistic regression\u003c/p\u003e\n\u003cp\u003eKNN \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;K nearest neighbors\u003c/p\u003e\n\u003cp\u003eSVM \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Support Vector Machines\u003c/p\u003e\n\u003cp\u003eXGB/XGBoost \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp;Extreme Gradient Boosting\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBT \u0026nbsp; \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Bagging tree algorithm\u003c/p\u003e\n\u003cp\u003eAda/AdaBoost\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Adaptive boosting\u003c/p\u003e\n\u003cp\u003eHP \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Hip fracture\u003c/p\u003e\n\u003cp\u003eMOF \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Major Osteoporotic Fracture Score\u003c/p\u003e\n\u003cp\u003eTBil \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Total bilirubin\u003c/p\u003e\n\u003cp\u003e\u0026beta;-CTx\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026beta;-Isomerized C-telopeptide\u003c/p\u003e\n\u003cp\u003eLH \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Luteinizing hormone\u003c/p\u003e\n\u003cp\u003eTP \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Total protein\u003c/p\u003e\n\u003cp\u003eWOMAC \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Osteoarthritis Index Score\u003c/p\u003e\n\u003cp\u003eAST \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Aspartate aminotransferase\u003c/p\u003e\n\u003cp\u003eALT \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Alanine aminotransferase\u003c/p\u003e\n\u003cp\u003eOC \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Osteocalcin\u003c/p\u003e\n\u003cp\u003eSCr \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Serum creatinine\u003c/p\u003e\n\u003cp\u003eT \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Testosterone\u003c/p\u003e\n\u003cp\u003eALP \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Alkaline phosphatase\u003c/p\u003e\n\u003cp\u003eCa \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Calcium\u003c/p\u003e\n\u003cp\u003eRBC \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Red blood cell count\u003c/p\u003e\n\u003cp\u003e\u0026gamma;-GT \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u0026gamma;-Glutamyl transpeptidase\u003c/p\u003e\n\u003cp\u003eHGH \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Growth hormone\u003c/p\u003e\n\u003cp\u003e25(OH)D \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;25-Hydroxy vitamin D\u003c/p\u003e\n\u003cp\u003eFSH \u0026nbsp;\u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;\u0026nbsp;Follicle-stimulating hormone\u003c/p\u003e\n\u003cp\u003ePRL \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp;Prolactin\u003c/p\u003e\n\u003cp\u003eBUN \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; \u0026nbsp; Blood urea nitrogen\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors would like to thank the patients who participated in this study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMeiling Chu and Guanghua Yang conceived and designed the experiments. Xiaowen Lai collected and analyzed the data,\u0026nbsp;Guanghua Yang and Lifang Song evaluated and verified the data and results again,\u0026nbsp;and\u0026nbsp;Meiling Chu and Xiaowen Lai wrote the paper. All authors have read and agreed to the published version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe following funds and projects supported this study: 1. Shanghai Seventh People\u0026apos;s Hospital Medical-Industrial Cross-Innovation Special Project (QYYGJ0102); 2. The Young Medical Talents Training Program of Shanghai Pudong New Area Health Commission (PWRq2024-46).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDue to patient privacy concerns within the Seventh People\u0026apos;s Hospital Affiliated with Shanghai University of Traditional Chinese Medicine Health Information Network, the primary data supporting this article cannot be shared publicly.Nevertheless, the data sets that were used in the current study are available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eEthics Approval and Consent to Participate This study was approved by the Institutional Review Board of the Seventh People\u0026apos;s Hospital Affiliated with Shanghai University of Traditional Chinese Medicine (approval number: ChiCTR2200057785). The requirement for informed consent was waived by the ethics committee due to the study\u0026rsquo;s retrospective nature. All procedures were conducted in accordance with relevant laws, regulations, and the principles outlined in the Declaration of HelsinkiConsent to participate The Institutional Review Board (IRB) of the Seventh People\u0026apos;s Hospital Affiliated with Shanghai University of Traditional Chinese Medicine waived the requirement for informed consent due to the retrospective nature of the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor details\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e1\u003c/sup\u003ethe Seventh People\u0026apos;s Hospital Affiliated with Shanghai University of Traditional Chinese Medicine,\u0026nbsp;No. 358, Datong Road, Shanghai 200137, Shanghai, China, Phone:+86 021-58670561.\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e\u0026dagger;\u003c/sup\u003eMeiling Chuand Xiaowen Lai contributed equally to this work as co-first authors.\u003c/p\u003e\n\u003cp\u003e\u003csup\u003e*\u003c/sup\u003eGuanghua Yangand Lifang Song contributed equally to this work as co-corresponding authors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCo-corresponding author\u003c/strong\u003e\u003cstrong\u003e：\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGuanghua Yang, the Seventh People\u0026apos;s Hospital Affiliated with Shanghai University of Traditional Chinese Medicine,\u0026nbsp;No. 358, Datong Road, Shanghai 200137, Shanghai, China, Phone:+86 021-58670561.\u003c/p\u003e\n\u003cp\u003eE-mail: [email protected]\u003c/p\u003e\n\u003cp\u003eLifang Song, the Seventh People\u0026apos;s Hospital Affiliated with Shanghai University of Traditional Chinese Medicine,\u0026nbsp;No. 358, Datong Road, Shanghai 200137, Shanghai, China, Phone:+86 021-58670561.\u003c/p\u003e\n\u003cp\u003eE-mail: [email protected]\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBRAY F, LAVERSANNE M, SUNG H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries[J]. CA: a cancer journal for clinicians, 2024, 74(3): 229-263. https://doi.org/10.3322/caac.21834.\u003c/li\u003e\n\u003cli\u003eXU Y, GONG M, WANG Y, et al. Global trends and forecasts of breast cancer incidence and deaths[J]. Scientific Data, 2023, 10: 334.https://doi.org/10.1038/s41597-023-02253-5.\u003c/li\u003e\n\u003cli\u003eBreast Cancer Expert Committee of the National Center for Quality Control in Oncology. The standardization of the whole-cycle management of breast cancer bone health[J]. Chinese Medical Journal, 2024, 104(2): 107-124.https://doi.org/10.3760/cma.j.cn112137-20231109-01132.\u003c/li\u003e\n\u003cli\u003eBreast Cancer Committee of the Chinese Anti-Cancer Association, Breast Cancer Group of the Oncology Branch of the Chinese Medical Association. Breast Cancer Diagnosis and Treatment Guidelines of the Chinese Anti-Cancer Association (2024 Edition)[J]. Chinese Journal of Cancer, 2023, 33(12): 1092-1187.https://doi.org/10.19401/j.cnki.1007-3639.2023.12.004.\u003c/li\u003e\n\u003cli\u003eLi H. Clinical study on the occurrence of osteoporosis associated with breast cancer chemotherapy [D]. Chongqing Medical University, 2020[2024-12-27].https://doi.org/10.27674/d.cnki.gcyku.2020.000226.\u003c/li\u003e\n\u003cli\u003eFATIMA N, LIU L, HONG S, et al. Prediction of Breast Cancer, Comparative Review of Machine Learning Techniques, and Their Analysis[J]. IEEE Access, 2020, 8: 150360-150376.https://doi.org/10.1109/ACCESS.2020.3016715.\u003c/li\u003e\n\u003cli\u003eTAKEFUJI Y. Evaluating feature importance biases in logistic regression: Recommendations for robust statistical methods[J]. European Journal of Internal Medicine, 2024: S0953620524004874.https://doi.org/10.1016/j.ejim.2024.11.022.\u003c/li\u003e\n\u003cli\u003eCAO T, WANG X, ZHANG H. Energy bagging tree[J]. Statistics and its interface, 2016, 9(2): 171-181.https://doi.org/10.4310/SII.2016.v9.n2.a5.\u003c/li\u003e\n\u003cli\u003eRIGATTI S J. Random Forest[J]. Journal of Insurance Medicine (New York, N.Y.), 2017, 47(1): 31-39.https://doi.org/10.17849/insm-47-01-31-39.1.\u003c/li\u003e\n\u003cli\u003eSARKER I H. Machine Learning: Algorithms, Real-World Applications and Research Directions[J]. Sn Computer Science, 2021, 2(3):160.https://doi.org/10.1007/s42979-021-00592-x.\u003c/li\u003e\n\u003cli\u003eCHEN T, GUESTRIN C. XGBoost: A Scalable Tree Boosting System[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, NY, USA: Association for Computing Machinery, 2016: 785-794[2024-12-26].https://doi.org/10.1145/2939672.2939785.\u003c/li\u003e\n\u003cli\u003eUDDIN S, HAQUE I, LU H, et al. Comparative performance analysis of K-nearest neighbour (KNN) algorithm and its different variants for disease prediction[J]. Scientific Reports, 2022, 12: 6256. https://doi.org/10.1038/s41598-022-10358-x.\u003c/li\u003e\n\u003cli\u003eBATTINENI G, CHINTALAPUDI N, AMENTA F. Machine learning in medicine: Performance calculation of dementia prediction by support vector machines (SVM)[J]. Informatics in Medicine Unlocked, 2019, 16: 100200.https://doi.org/10.1016/j.imu.2019.100200.\u003c/li\u003e\n\u003cli\u003eTANG D D, YE Z J, LIU W W, et al. Survival feature and trend of female breast cancer: A comprehensive review of survival analysis from cancer registration data[J]. The Breast, 2025, 79[2024-12-26].https://doi.org/10.1016/j.breast.2024.103862.\u003c/li\u003e\n\u003cli\u003eMUGNIER B, GONCALVES A, DAUMAS A, et al. Prevention of aromatase inhibitor\u0026ndash;induced bone loss with anti-resorptive therapy in post-menopausal women with early-stage breast cancer[J]. Osteoporosis International, 2023, 34(4): 703-711.https://doi.org/10.1007/s00198-023-06683-0.\u003c/li\u003e\n\u003cli\u003eCHENG L L, YE F, XU T, et al. Nomogram Model for Predicting Minimal Breast Cancer Based on Clinical and Ultrasonic Characteristics[J]. International Journal of Women\u0026rsquo;s Health, 2024, 16: 2173-2184. https://doi.org/10.2147/IJWH.S482291.\u003c/li\u003e\n\u003cli\u003eWEN X, CHEN J, ZHONG J, et al. Model based on ultrasound and clinicopathological characteristics for early prediction of pathological complete response to neoadjuvant chemotherapy in breast cancer[J]. Quantitative Imaging in Medicine and Surgery, 2024, 14(12): 8840-8851.https://doi.org/10.21037/qims-24-1268.\u003c/li\u003e\n\u003cli\u003eSAPIENZA L G, SALCEDO M P, NING M S, et al. Pelvic Insufficiency Fractures After External Beam Radiation Therapy for Gynecologic Cancers: A Meta-analysis and Meta-regression of 3929 Patients[J]. International Journal of Radiation Oncology, Biology, Physics, 2020, 106(3): 475-484.https://doi.org/10.1016/j.ijrobp.2019.09.012.\u003c/li\u003e\n\u003cli\u003eMOHAMMADI S M, SANIEE N, MOUSAVIASL S, et al. The Role of Osteocalcin in Patients with Osteoporosis: A Systematic Review[J]. Iranian Journal of Public Health, 2024, 53(11): 2432-2439.https://doi.org/10.18502/ijph.v53i11.16945.\u003c/li\u003e\n\u003cli\u003eLEE J H, PARK J, KIM J H, et al. Integrative analysis of genetic and clinical risk factors for bone loss in a Korean population[J]. Bone, 2021, 147: 115910.https://doi.org/10.1016/j.bone.2021.115910.\u003c/li\u003e\n\u003cli\u003eNAGAYAMA Y, EBINA K, TSUBOI H, et al. Low serum albumin concentration is associated with increased risk of osteoporosis in postmenopausal patients with rheumatoid arthritis[J]. Journal of Orthopedic Science, 2022, 27(6): 1283-1290. https://doi.org/10.1016/j.jos.2021.08.018.\u003c/li\u003e\n\u003cli\u003eLEI S, ZHANG X, SONG L, et al. Expert consensus on vitamin D in osteoporosis[J]. Annals of Joint, 2025, 10: 1.https://doi.org/10.21037/aoj-24-48.\u003c/li\u003e\n\u003cli\u003eLI G H Y, ROBINSON-COHEN C, SAHNI S, et al. Association of Genetic Variants Related to Serum Calcium Levels with Reduced Bone Mineral Density[J]. The Journal of Clinical Endocrinology and Metabolism, 2019, 105(3): e328-e336. https://doi.org/10.1210/clinem/dgz088.\u003c/li\u003e\n\u003cli\u003ePENG W, LI Z, GUAN Y, et al. A study of cognitive functions in female elderly patients with osteoporosis: a multicenter cross-sectional study[J]. Aging \u0026amp; Mental Health, 2016, 20(6): 647-654.https://doi.org/10.1080/13607863.2015.1033680.\u003c/li\u003e\n\u003cli\u003eNAKAJIMA K, HORII C, KODAMA H, et al. Association between vertebral fractures and brain volume: insights from a community cohort study[J]. Osteoporosis International, 2025[2025-03-24].https://doi.org/10.1007/s00198-025-07403-6.\u003c/li\u003e\n\u003cli\u003eKIM S H, CHO Y U, KIM S J, et al. Changes in Bone Mineral Density in Women With Breast Cancer: A Prospective Cohort Study[J]. Cancer Nursing, 2019, 42(2): 164-172.https://doi.org/10.1097/NCC.0000000000000586.\u003c/li\u003e\n\u003cli\u003eWang ZG. Ling Shu Jing [M]. Beijing: China Press of Traditional Chinese Medicine, 2022: 156.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"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":"bmc-medical-informatics-and-decision-making","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"midm","sideBox":"Learn more about [BMC Medical Informatics and Decision Making](http://bmcmedinformdecismak.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/midm/default.aspx","title":"BMC Medical Informatics and Decision Making","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Breast cancer, Bone loss, Risk prediction model, Learning algorithm","lastPublishedDoi":"10.21203/rs.3.rs-6455887/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6455887/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eBreast cancer is among the most prevalent malignant neoplasms afflicting women globally, with its treatment and disease itself exerting a substantial influence on bone health. Given the prolonged survival rates of breast cancer patients, the management of bone health issues has become a critical component of comprehensive cancer care. The objective of this study was to develop clinical predictive models using machine learning methods and apply these models to the management of bone health in breast cancer patients.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eThis was a multicenter retrospective cohort study. We included 139 breast cancer patients who were diagnosed from March 2022 to December 2024 in three hospitals in China. We developed predictive models with optimal features by using algorithms such as random forest (RF), K nearest neighbors (KNN), support vector machines (SVM), and extreme gradient boosting (XGB) and determined and assessed the machine learning algorithm with the highest accuracy rate for breast cancer-related bone loss on the basis of the area under curve (AUC) of the subjects.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eA total of 139 study participants were included in this study, including 78 patients with osteopenia (including osteoporosis, T\u0026le;-1.0) and 61 patients with normal bone mass (T\u0026gt;-1.0). Specific indicators of bone loss were identified, and seven models were constructed: logistic regression (LR), the bagging tree algorithm (BT), RF, adaptive boosting (AdaBoost), XGBoost, KNN, and SVM, among which the AdaBoost prediction model performed the best in predicting breast cancer-related bone loss, with the best performance AUC of 0.995. The model can be generated into a publicly accessible applet.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003e The data was incorporated into seven machine learning algorithms and models were constructed, which were then compared to arrive at the optimal model.The AdaBoost model was selected as the final model, which can predict breast cancer-related bone loss by collecting patient information through a simple question-and-answer session and importing laboratory blood tests and reports of examinations such as bone mineral density; this model is expected to guide the management of clinical bone health in patients with breast cancer and improve the prognosis of patients.\u003c/p\u003e\u003ch2\u003eTrial registration\u003c/h2\u003e \u003cp\u003e Ethics Approval and Consent to Participate This study was approved by the Institutional Review Board of the Seventh People's Hospital Affiliated with Shanghai University of Traditional Chinese Medicine. Approval number: ChiCTR2200057785(17th March 2022).\u003c/p\u003e","manuscriptTitle":"A study on the development of a machine learning-based clinical prediction model for bone health management in breast cancer patients","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-08 14:45:49","doi":"10.21203/rs.3.rs-6455887/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-05-26T04:13:08+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-25T15:15:12+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"146392638592897428666816047459823326866","date":"2025-05-23T06:13:19+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"54264363322931605203270737550712069640","date":"2025-05-23T05:27:51+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-22T07:47:26+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"223017803365079518416671518383718573661","date":"2025-05-21T11:32:04+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-21T05:10:07+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"288489020536322633609613717595296073350","date":"2025-05-20T13:51:16+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"219077741148528420949790099337757700490","date":"2025-05-06T20:05:25+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"60747127356837615972079826410487920477","date":"2025-05-06T02:19:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"136124258696432240306646379902337842403","date":"2025-05-03T08:11:54+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"225168238298715832830659050221518495897","date":"2025-05-02T13:49:06+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-05-01T01:31:50+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-04-22T14:42:44+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-04-21T09:29:10+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-04-21T09:27:32+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Medical Informatics and Decision Making","date":"2025-04-15T14:31:38+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-medical-informatics-and-decision-making","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"midm","sideBox":"Learn more about [BMC Medical Informatics and Decision Making](http://bmcmedinformdecismak.biomedcentral.com/)","snPcode":"","submissionUrl":"https://www.editorialmanager.com/midm/default.aspx","title":"BMC Medical Informatics and Decision Making","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"em","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"6c993d47-2612-465a-be43-64907623a4a4","owner":[],"postedDate":"May 8th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[],"tags":[],"updatedAt":"2025-11-24T05:53:13+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-08 14:45:49","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6455887","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6455887","identity":"rs-6455887","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}