Tianjin Medical Journal ›› 2024, Vol. 52 ›› Issue (7): 775-780.doi: 10.11958/20231584
• Applied Research • Previous Articles Next Articles
WANG Xian1(), LIU Xiaming1, CHEN Manyu1, ZHAO Jun1, WANG Lidong2,∆(
)
Received:
2023-10-27
Revised:
2024-02-06
Published:
2024-07-15
Online:
2024-07-11
Contact:
∆E-mail:whlfdjs@126.com
WANG Xian, LIU Xiaming, CHEN Manyu, ZHAO Jun, WANG Lidong. Construction and verification of prediction model of type 2 diabetic nephropathy based on machine learning[J]. Tianjin Medical Journal, 2024, 52(7): 775-780.
CLC Number:
组别 | n | 男性 | 年龄/岁 | 高血压病 | 冠心病 | 脑血管病 | 痛风 | 糖尿病家族史 | 吸烟史 | 饮酒史 | ||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
非DKD组 | 281 | 145(51.6) | 56(49,63) | 129(45.9) | 27(9.6) | 27(9.6) | 10(3.6) | 167(59.4) | 103(36.7) | 112(39.9) | ||||||||||||||||||||||||
DKD组 | 89 | 53(59.6) | 58(51,65) | 57(64.0) | 7(7.9) | 7(7.9) | 7(7.9) | 59(66.3) | 40(44.9) | 40(44.9) | ||||||||||||||||||||||||
χ2或Z | 1.717 | 2.507* | 8.894** | 0.246 | 0.246 | 1.962 | 1.339 | 1.959 | 0.722 | |||||||||||||||||||||||||
组别 | BMI≥24.0 kg/m2 | 脂肪肝 | 糖尿病病程 | VFA/cm2 | DR | HbA1c/% | ||||||||||||||||||||||||||||
<1年 | 1~<10年 | ≥10年 | NDR | NPDR | PDR | |||||||||||||||||||||||||||||
非DKD组 | 200(71.2) | 181(64.4) | 23(8.2) | 131(46.6) | 127(45.2) | 93(73,118) | 193(68.7) | 86(30.6) | 2(0.7) | 8.9(7.6,10.5) | ||||||||||||||||||||||||
DKD组 | 76(85.4) | 81(91.0) | 4(4.5) | 29(32.6) | 56(62.9) | 102(84,122) | 34(38.2) | 43(48.3) | 12(13.5) | 8.9(7.9,10.6) | ||||||||||||||||||||||||
χ2或Z | 7.211** | 23.136** | 2.913** | 1.924 | 5.716** | 0.692 | ||||||||||||||||||||||||||||
组别 | FC-P/(mmol/L) | ALT/(U/L) | AST/(U/L) | UA/(μmol/L) | CR/(μmol/L) | TG/(mmol/L) | TC/(mmol/L) | HDL-C/(mmol/L) | ||||||||||||||||||||||||||
非DKD组 | 1.87(1.29,2.64) | 19(14,28) | 16(13,22) | 279±82 | 53(43,62) | 1.78(1.24,2.70) | 4.5(3.8,5.3) | 1.20(1.05,1.34) | ||||||||||||||||||||||||||
DKD组 | 1.57(1.16,2.45) | 15(11,22) | 14(12,17) | 301±99 | 56(47,75) | 2.19(1.31,3.54) | 4.3(3.6,5.5) | 1.17(0.99,1.36) | ||||||||||||||||||||||||||
Z或t | 1.826 | 3.631** | 3.150** | 2.136* | 3.325** | 2.312* | 0.390 | 0.926 | ||||||||||||||||||||||||||
组别 | LDL-C/(mmol/L) | Cys-C/(mg/L) | PTH/(ng/L) | 25(OH)D/(μg/L) | Lym/(×109/L) | Mon/(×109/L) | TyG | |||||||||||||||||||||||||||
非DKD组 | 2.39±0.79 | 0.95(0.86,1.05) | 36.63(28.69,44.66) | 14.88(10.32,19.06) | 1.97(1.59,2.53) | 0.37(0.29,0.47) | 9.34±0.79 | |||||||||||||||||||||||||||
DKD组 | 2.28±1.01 | 1.07(0.96,1.21) | 33.27(22.94,44.13) | 11.73(8.64,16.06) | 2.05(1.38,2.78) | 0.44(0.31,0.63) | 9.58±0.98 | |||||||||||||||||||||||||||
t或Z | 1.097 | 5.728** | 1.947 | 3.764** | 0.020 | 2.658** | 2.129* |
Tab.1 Comparison of clinical data between the two groups
组别 | n | 男性 | 年龄/岁 | 高血压病 | 冠心病 | 脑血管病 | 痛风 | 糖尿病家族史 | 吸烟史 | 饮酒史 | ||||||||||||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
非DKD组 | 281 | 145(51.6) | 56(49,63) | 129(45.9) | 27(9.6) | 27(9.6) | 10(3.6) | 167(59.4) | 103(36.7) | 112(39.9) | ||||||||||||||||||||||||
DKD组 | 89 | 53(59.6) | 58(51,65) | 57(64.0) | 7(7.9) | 7(7.9) | 7(7.9) | 59(66.3) | 40(44.9) | 40(44.9) | ||||||||||||||||||||||||
χ2或Z | 1.717 | 2.507* | 8.894** | 0.246 | 0.246 | 1.962 | 1.339 | 1.959 | 0.722 | |||||||||||||||||||||||||
组别 | BMI≥24.0 kg/m2 | 脂肪肝 | 糖尿病病程 | VFA/cm2 | DR | HbA1c/% | ||||||||||||||||||||||||||||
<1年 | 1~<10年 | ≥10年 | NDR | NPDR | PDR | |||||||||||||||||||||||||||||
非DKD组 | 200(71.2) | 181(64.4) | 23(8.2) | 131(46.6) | 127(45.2) | 93(73,118) | 193(68.7) | 86(30.6) | 2(0.7) | 8.9(7.6,10.5) | ||||||||||||||||||||||||
DKD组 | 76(85.4) | 81(91.0) | 4(4.5) | 29(32.6) | 56(62.9) | 102(84,122) | 34(38.2) | 43(48.3) | 12(13.5) | 8.9(7.9,10.6) | ||||||||||||||||||||||||
χ2或Z | 7.211** | 23.136** | 2.913** | 1.924 | 5.716** | 0.692 | ||||||||||||||||||||||||||||
组别 | FC-P/(mmol/L) | ALT/(U/L) | AST/(U/L) | UA/(μmol/L) | CR/(μmol/L) | TG/(mmol/L) | TC/(mmol/L) | HDL-C/(mmol/L) | ||||||||||||||||||||||||||
非DKD组 | 1.87(1.29,2.64) | 19(14,28) | 16(13,22) | 279±82 | 53(43,62) | 1.78(1.24,2.70) | 4.5(3.8,5.3) | 1.20(1.05,1.34) | ||||||||||||||||||||||||||
DKD组 | 1.57(1.16,2.45) | 15(11,22) | 14(12,17) | 301±99 | 56(47,75) | 2.19(1.31,3.54) | 4.3(3.6,5.5) | 1.17(0.99,1.36) | ||||||||||||||||||||||||||
Z或t | 1.826 | 3.631** | 3.150** | 2.136* | 3.325** | 2.312* | 0.390 | 0.926 | ||||||||||||||||||||||||||
组别 | LDL-C/(mmol/L) | Cys-C/(mg/L) | PTH/(ng/L) | 25(OH)D/(μg/L) | Lym/(×109/L) | Mon/(×109/L) | TyG | |||||||||||||||||||||||||||
非DKD组 | 2.39±0.79 | 0.95(0.86,1.05) | 36.63(28.69,44.66) | 14.88(10.32,19.06) | 1.97(1.59,2.53) | 0.37(0.29,0.47) | 9.34±0.79 | |||||||||||||||||||||||||||
DKD组 | 2.28±1.01 | 1.07(0.96,1.21) | 33.27(22.94,44.13) | 11.73(8.64,16.06) | 2.05(1.38,2.78) | 0.44(0.31,0.63) | 9.58±0.98 | |||||||||||||||||||||||||||
t或Z | 1.097 | 5.728** | 1.947 | 3.764** | 0.020 | 2.658** | 2.129* |
数据集 | LR | KNN | SVM | |||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
AUC(95%CI) | 敏感度 | 特异度 | AUC(95%CI) | 敏感度 | 特异度 | AUC(95%CI) | 敏感度 | 特异度 | ||||||||||
训练集 | 0.786(0.730~0.842) | 0.326 | 0.950 | 0.981(0.970~0.991) | 0.618 | 0.986 | 0.871(0.824~0.918) | 0.438 | 0.979 | |||||||||
验证集 | 0.773(0.691~0.855) | 0.316 | 0.950 | 0.793(0.711~0.875) | 0.368 | 0.933 | 0.776(0.685~0.867) | 0.342 | 0.967 | |||||||||
D | 0.255 | 4.467** | 1.815 | |||||||||||||||
数据集 | DT | RF | NB | |||||||||||||||
AUC(95%CI) | 敏感度 | 特异度 | AUC(95%CI) | 敏感度 | 特异度 | AUC(95%CI) | 敏感度 | 特异度 | ||||||||||
训练集 | 0.850(0.799~0.901) | 0.663 | 0.929 | 1.000(0.999~1.000) | 0.865 | 1.000 | 0.739(0.678~0.801) | 0.348 | 0.932 | |||||||||
验证集 | 0.717(0.625~0.809) | 0.368 | 0.892 | 0.777(0.694~0.860) | 0.316 | 0.917 | 0.718(0.620~0.817) | 0.316 | 0.942 | |||||||||
D | 2.475* | 5.264** | 0.351 | |||||||||||||||
数据集 | ANN | XGBoost | ||||||||||||||||
AUC(95%CI) | 敏感度 | 特异度 | AUC(95%CI) | 敏感度 | 特异度 | |||||||||||||
训练集 | 0.835(0.788~0.882) | 0.697 | 0.854 | 0.918(0.887~0.949) | 0.730 | 0.904 | ||||||||||||
验证集 | 0.738(0.640~0.836) | 0.632 | 0.800 | 0.709(0.616~0.802) | 0.289 | 0.892 | ||||||||||||
D | 1.753 | 4.175** |
Tab.2 Comparison of predictive performance of ML models in training set and validation set
数据集 | LR | KNN | SVM | |||||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
AUC(95%CI) | 敏感度 | 特异度 | AUC(95%CI) | 敏感度 | 特异度 | AUC(95%CI) | 敏感度 | 特异度 | ||||||||||
训练集 | 0.786(0.730~0.842) | 0.326 | 0.950 | 0.981(0.970~0.991) | 0.618 | 0.986 | 0.871(0.824~0.918) | 0.438 | 0.979 | |||||||||
验证集 | 0.773(0.691~0.855) | 0.316 | 0.950 | 0.793(0.711~0.875) | 0.368 | 0.933 | 0.776(0.685~0.867) | 0.342 | 0.967 | |||||||||
D | 0.255 | 4.467** | 1.815 | |||||||||||||||
数据集 | DT | RF | NB | |||||||||||||||
AUC(95%CI) | 敏感度 | 特异度 | AUC(95%CI) | 敏感度 | 特异度 | AUC(95%CI) | 敏感度 | 特异度 | ||||||||||
训练集 | 0.850(0.799~0.901) | 0.663 | 0.929 | 1.000(0.999~1.000) | 0.865 | 1.000 | 0.739(0.678~0.801) | 0.348 | 0.932 | |||||||||
验证集 | 0.717(0.625~0.809) | 0.368 | 0.892 | 0.777(0.694~0.860) | 0.316 | 0.917 | 0.718(0.620~0.817) | 0.316 | 0.942 | |||||||||
D | 2.475* | 5.264** | 0.351 | |||||||||||||||
数据集 | ANN | XGBoost | ||||||||||||||||
AUC(95%CI) | 敏感度 | 特异度 | AUC(95%CI) | 敏感度 | 特异度 | |||||||||||||
训练集 | 0.835(0.788~0.882) | 0.697 | 0.854 | 0.918(0.887~0.949) | 0.730 | 0.904 | ||||||||||||
验证集 | 0.738(0.640~0.836) | 0.632 | 0.800 | 0.709(0.616~0.802) | 0.289 | 0.892 | ||||||||||||
D | 1.753 | 4.175** |
SVM | LR | NB | ANN | |||||
---|---|---|---|---|---|---|---|---|
训练集 | 验证集 | 训练集 | 验证集 | 训练集 | 验证集 | |||
训练集 | 3.105** | 4.467** | 1.221 | |||||
验证集 | 0.075 | 1.309 | 1.309 |
Tab.3 Delong test between SVM model with LR, NB, and ANN model
SVM | LR | NB | ANN | |||||
---|---|---|---|---|---|---|---|---|
训练集 | 验证集 | 训练集 | 验证集 | 训练集 | 验证集 | |||
训练集 | 3.105** | 4.467** | 1.221 | |||||
验证集 | 0.075 | 1.309 | 1.309 |
[1] | VARGHESE R T, JIALAL I. Diabetic Nephropathy[M]. StatPearls. Treasure Island(FL): StatPearls Publishing Copyright© 2023,StatPearls Publishing LLC,2023. |
[2] | ZHAO Z H, HUO L L, WANG L Y, et al. Survival of Chinese people with type 2 diabetes and diabetic kidney disease:a cohort of 12-year follow-up[J]. BMC Public Health, 2019, 19(1):1498. doi:10.1186/s12889-019-7859-x. |
[3] | YUE S R, LI S S, HUANG X Y, et al. Machine learning for the prediction of acute kidney injury in patients with sepsis[J]. J Transl Med, 2022, 20(1):215. doi:10.1186/s12967-022-03364-0. |
[4] | ABEGAZ T M, BALJOON A, KILANKO O, et al. Machine learning algorithms to predict major adverse cardiovascular events in patients with diabetes[J]. Comput Biol Med, 2023,164:107289. doi:10.1016/j.compbiomed.2023.107289. |
[5] | 中华医学会糖尿病学分会. 中国2型糖尿病防治指南(2020年版)[J]. 中华糖尿病杂志, 2021, 13(4):315-409. |
Chinese Diabetes Society. Guideline for the prevention and treatment of type 2 diabetes mellitus in China(2020 edition)[J]. Chin J Diabetes Mellitus, 2021, 13(4):315-409. doi:10.3760/cma.j.cn115791-20210221-00095. | |
[6] | 中华医学会糖尿病学分会微血管并发症学组. 中国糖尿病肾脏病防治指南(2021年版)[J]. 中华糖尿病杂志, 2021, 13(8):762-784. |
Microvascular Complications Group of Chinese Diabetes Society. Clinical guideline for the prevention and treatment of diabetic kidney disease in China(2021 edition)[J]. Chin J Diabetes Mellitus, 2021, 13(8):762-784. doi:10.3760/cma.j.cn115791-20210706-00369. | |
[7] | PERKINS B A, BEBU I, DE BOER I H, et al. Risk factors for kidney disease in type 1 diabetes[J]. Diabetes Care, 2019, 42(5):883-890. doi:10.2337/dc18-2062. |
[8] | XU B, LI S Q, KANG B, et al. The current role of sodium-glucose cotransporter 2 inhibitors in type 2 diabetes mellitus management[J]. Cardiovasc Diabetol, 2022, 21(1):83. doi:10.1186/s12933-022-01512-w. |
[9] | ZOU Y T, ZHAO L J, ZHANG J L, et al. Development and internal validation of machine learning algorithms for end-stage renal disease risk prediction model of people with type 2 diabetes mellitus and diabetic kidney disease[J]. Ren Fail, 2022, 44(1):562-570. doi:10.1080/0886022x.2022.2056053. |
[10] | LIU W, DU J, GE X X, et al. The analysis of risk factors for diabetic kidney disease progression:a single-centre and cross-sectional experiment in Shanghai[J]. BMJ Open, 2022, 12(6):e060238. doi:10.1136/bmjopen-2021-060238. |
[11] | YANAGAWA T, KOYANO K, AZUMA K. Retrospective study of factors associated with progression and remission/regression of diabetic kidney disease-hypomagnesemia was associated with progression and elevated serum alanine aminotransferase levels were associated with remission or regression[J]. Diabetol Int, 2021, 12(3):268-276. doi:10.1007/s13340-020-00483-1. |
[12] | PETROFF D, BÄTZ O, JEDRYSIAK K, et al. Age dependence of liver enzymes:an analysis of over 1,300,000 consecutive blood samples[J]. Clin Gastroenterol Hepatology, 2022, 20(3):641-650. doi:10.1016/j.cgh.2021.01.039. |
[13] | FELÍCIO J S, DE RIDER BRITTO H A, CORTEZ P C, et al. Association between 25(OH)vitamin D,HbA1c and albuminuria in diabetes mellitus:data from a population-based study (VIDAMAZON)[J]. Front Endocrinol (Lausanne), 2021,12:12723502. doi:10.3389/fendo.2021.723502. |
[14] | XU F, LU H Y, LAI T W, et al. Association between vitamin D status and mortality among adults with diabetic kidney disease[J]. J Diabetes Res, 2022,2022:9632355. doi:10.1155/2022/9632355. |
[15] | HUANG H Y, LIN T W, HONG Z X, et al. Vitamin D and diabetic kidney disease[J]. Int J Mol Sci, 2023, 24(4):3571. doi:10.3390/ijms24043751. |
[16] | NISHAD R, MUKHI D, KETHAVATH S, et al. Podocyte derived TNF-α mediates monocyte differentiation and contributes to glomerular injury[J]. FASEB J, 2022, 36(12):e22622. doi:10.1096/fj.202200923R. |
[17] | 许莉敏, 谢燕. 外周血单核细胞DNMT1及血清IL-6在糖尿病肾脏病中的表达及意义[J]. 天津医药, 2023, 51(2):194-198. |
XU L M, XIE Y. Expression and significance of peripheral blood mononuclear cell DNMT1 and serum IL-6 in diabetic nephropathy[J]. Tianjin Med J, 2023, 51(2):194-198. doi:10.11958/20220812. | |
[18] | ZENG H X, QI X M, XU X X, et al. TAB1 regulates glycolysis and activation of macrophages in diabetic nephropathy[J]. Inflamm Res, 2020, 69(12):1215-1234. doi:10.1007/s00011-020-01411-4. |
[19] | OPAZO-RÍOS L, MAS S, MARÍN-ROYO G, et al. Lipotoxicity and diabetic nephropathy: novel mechanistic insights and therapeutic opportunities[J]. Int J Mol Sci, 2020, 21(7):2632. doi:10.3390/ijms21072632. |
[20] | YU K P, LI D R, XU F P, et al. IDO1 as a new immune biomarker for diabetic nephropathy and its correlation with immune cell infiltration[J]. Int Immunopharmacol, 2021,94:107446. doi:10.1016/j.intimp.2021.107446. |
Viewed | ||||||
Full text |
|
|||||
Abstract |
|
|||||