基于机器学习算法的老年人慢性肾脏病预测模型
PDF下载 (55)张 良,冯 伟,李 宁,纪 威,许国章.基于机器学习算法的老年人慢性肾脏病预测模型[J].宁波大学学报(理工版),2026,39(2):53-58.DOI:10.20098/j.cnki.1001-5132.2025.0722
ZHANG Liang,FENG Wei,LI Ning,JI Wei,XU Guozhang.Prediction model for chronic kidney disease in the elderly people based on machine learning[J].Journal of Ningbo University(Natural Science & Engineering Edition),2026,39(2):53-58.DOI:10.20098/j.cnki.1001-5132.2025.0722
| Title: | Prediction model for chronic kidney disease in the elderly people based on machine learning |
| 作者: | 张 良, 冯 伟, 李 宁, 纪 威, 许国章 |
| Author(s): | ZHANG Liang, FENG Wei, LI Ning, JI Wei, XU Guozhang |
| 关键词: | 机器学习; 梯度提升机; 慢性肾脏病; 预测模型 |
| Keywords: | machine learning; gradient boosting machine; chronic kidney disease; prediction model |
| 分类号: | R692 |
| DOI: | 10.20098/j.cnki.1001-5132.2025.0722 |
| 文献标识码: | A |
| 摘要: | 采用机器学习算法构建老年人慢性肾脏病预测模型，为早期发现老年人慢性肾脏病高风险人群提供理论依据。选取宁波市某区82196名老年人作为研究对象，使用R 4.0.5软件构建Logistic回归、分类回归树、梯度提升机和支持向量机4种机器学习算法的老年人慢性肾脏病风险预测模型，并采用准确度、灵敏度、特异度、受试者工作曲线（ROC）及曲线下面积（AUC）来评估4种机器学习算法的预测效果。结果表明，共有40585名纳入本研究的老年人新发慢性肾脏病的发病率为5.43%；多因素Logistic回归分析结果显示，年龄、体质指数、腰围、糖尿病、血脂异常与老年人慢性肾脏病呈正相关（P<0.05）；经4种算法比较后，发现梯度提升机在训练集和验证集中的AUC均为最高，其值分别为0.939和0.941，表明其综合预测性能最好。 |
| Abstract: | This paper aims to construct a prediction model for Chronic Kidney Disease (CKD) in the elderly by using machine learning algorithms, so as to provide a basis for the early detection of high-risk groups of CKD in these people. A total of 82196 elderly individuals in a certain district of Ningbo City were selected as the research subjects. Four machine learning algorithms, namely logistic regression, Classification and Regression Tree (CART), Gradient Boosting Machine (GBM), and Support Vector Machine (SVM), were used to construct CKD risk prediction models for the elderly using R 4.0.5. The prediction performance of the four machine learning algorithms was evaluated using accuracy, sensitivity, specificity, Receiver Operating Characteristic (ROC) curve, and Area Under the Curve (AUC). A total of 40585 elderly individuals included in this study got new-onset CKD, and its incidence in the elderly was 5.43%. The multivariate logistic regression analysis showed that age, body mass index, waist circumference, diabetes, and dyslipidemia were positively correlated with CKD in the elderly (P<0.05). Among the four algorithms, the Gradient Boosting Machine (GBM) had the highest AUC of 0.939 in the training set and 0.941 in the validation set, demonstrating the best overall prediction performance. |
| 参考文献 /References: | [1] 上海市肾内科临床质量控制中心专家组. 慢性肾脏病早期筛查、诊断及防治指南(2022年版)[J]. 中华肾脏病杂志, 2022, 38(5):453-464. [2] GUO J X, LIU Z, WANG P J, et al. Global, regional, and national burden inequality of chronic kidney disease, 1990—2021: a systematic analysis for the global burden of disease study 2021[J]. Frontiers in medicine, 2025, 11: 1501175. [3] 卞帅博, 张亮, 杨璇. 1990—2021年中国慢性肾脏病疾病负担变化趋势及发病预测分析[J]. 实用临床医药杂志, 2025, 29(6):89-93, 98. [4] LI Y, NING Y, SHEN B, et al. Temporal trends in prevalence and mortality for chronic kidney disease in China from 1990 to 2019: an analysis of the global burden of disease study 2019[J]. Clinical kidney journal, 2022, 16(2):312-321. [5] LAFFIN L J, BAKRIS G L. Intersection between chronic kidney disease and cardiovascular disease[J]. Current cardiology reports, 2021, 23(9):117. [6] MATSUSHITA K, BALLEW S H, WANG A Y, et al. Epidemiology and risk of cardiovascular disease in populations with chronic kidney disease[J]. Nature reviews nephrology, 2022, 18(11):696-707. [7] 李婷婷, 苏鹏, 陈晋波, 等. 基于机器学习比较2型糖尿病患者发生糖尿病肾脏疾病风险预测模型的研究[J]. 中国糖尿病杂志, 2025, 33(4):241-247. [8] 王淑玉, 杜红阳, 赵晨昊, 等. 基于深度学习和机器学习的胃癌预测模型构建及评估[J]. 青岛大学学报(医学版), 2025, 61(1):54-58. [9] DELGADO C, BAWEJA M, CREWS D C, et al. A unifying approach for GFR estimation: recommendations of the NKF-ASN task force on reassessing the inclusion of race in diagnosing kidney disease[J]. American journal of kidney diseases, 2022, 79(2):268-288. [10] 诸骏仁, 高润霖, 赵水平, 等. 中国成人血脂异常防治指南(2016年修订版)[J]. 中国循环杂志, 2016, 31(10): 937-953. [11] LUO W, LEI L, LAI J C, et al. A risk prediction model for new-onset chronic kidney disease in the elderly[J]. American journal of nephrology, 2025, 56(1):58-69. [12] JOO Y S, KIM H W, LEE S M, et al. Dietary zinc intake and incident chronic kidney disease[J]. Clinical nutrition, 2021, 40(3):1039-1045. [13] GUPPY M, THOMAS E T, GLASZIOU P, et al. Rate of decline in kidney function with age: a systematic review[J]. BMJ open, 2024, 14(11):e089783. [14] JIANG Y, CAO Q, HONG W Q, et al. Age and estimated glomerular filtration rate in Chinese older adults: a cohort study from 2014 to 2020[J]. Frontiers in public health, 2024,12:1392903. [15] EBERT N, JAKOB O, GAEDEKE J, et al. Prevalence of reduced kidney function and albuminuria in older adults: the Berlin initiative study[J]. Nephrology dialysis trans- plantation, 2017, 32(6):997-1005. [16] LIU P, QUINN R R, LAM N N, et al. Accounting for age in the definition of chronic kidney disease[J]. JAMA internal medicine, 2021, 181(10):1359-1366. [17] WANG Y, CHEN X, SONG Y, et al. Association between obesity and kidney disease: a systematic review and meta- analysis[J]. Kidney international, 2008, 73(1):19-33. [18] CHOI J R, KOH S B, CHOI E. Waist-to-height ratio index for predicting incidences of hypertension: the ARIRANG study[J]. BMC public health, 2018, 18(1):767. [19] JITRAKNATEE J, RUENGORN C, NOCHAIWONG S. Prevalence and risk factors of chronic kidney disease among type 2 diabetes patients: a cross-sectional study in primary care practice[J]. Scientific reports, 2020, 10(1): 6205. [20] JOSHI R, SUBEDI P, YADAV G K, et al. Prevalence and risk factors of chronic kidney disease among patients with type 2 diabetes mellitus at a tertiary care hospital in Nepal: a cross-sectional study[J]. BMJ open, 2023, 13(2):e067238. [21] KIRKMAN M S, MAHMUD H, KORYTKOWSKI M T. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes[J]. Endocrinology and metabolism clinics of north America, 2018, 47(1):81-96. [22] SINHA S K, NICHOLAS S B. Pathomechanisms of diabetic kidney disease[J]. Journal of clinical medicine, 2023, 12(23):7349. [23] WANG M Z, XIA M, YANG H Y, et al. Interaction effect of blood glucose and pressure on the risk of chronic kidney disease: a population-based prospective cohort study[J]. Endocrine, 2022, 77(2):252-261. [24] CUNILLERA-PUÉRTOLAS O, VIZCAYA D, COBO- GUERRERO S, et al. Chronic kidney disease progression in patients with previous type 2 diabetes and/or hyper- tension: a population-based cohort study from primary care in Spain[J]. BMJ open, 2025, 15(1):e086919. [25] RUSSO G T, DE COSMO S, VIAZZI F, et al. Plasma triglycerides and HDL-C levels predict the development of diabetic kidney disease in subjects with type 2 diabetes: the AMD annals initiative[J]. Diabetes care, 2016, 39(12): 2278-2287. [26] MIGDALIS I N, IOANNIDIS I M, PAPANAS N, et al. Hypertriglyceridemia and other risk factors of chronic kidney disease in type 2 diabetes: a hospital-based clinic population in Greece[J]. Journal of clinical medicine, 2022, 11(11):3224. [27] SONG W Z, LIU Y F, QIU L X, et al. Machine learning- based warning model for chronic kidney disease in individuals over 40 years old in underprivileged areas, Shanxi province[J]. Frontiers in medicine, 2023, 9:930541. |
| 备注/Memo: | 收稿日期：2025-07-29 宁波大学学报（理工版）网址：http://journallg.nbu.edu.cn/ 基金项目：国家自然科学基金（72074011）；宁波市医疗卫生高端团队重大攻坚项目（2023020713） 第一作者：张 良，高级工程师，主要研究方向为信息化建设与大数据。E-mail: johnllen@qq.com *通信作者：许国章，博士/教授，主要研究方向为现场流行病学。E-mail: xuguozhang@nbu.edu.cn 宁波大学学报（理工版）网址：http://journallg.nbu.edu.cn/ |
