Influencing factors of occurrence of gestational diabetes mellitus and effectiveness analysis of Lasso-Nomogram prediction model
Objective To analyze the pathogenic factors of gestational diabetes mellitus(GDM),and build a Lasso-Nomogram predic-tion model.Methods A total of 150 pregnant women were selected from Hangzhou Women's Hospital from June 2020 to January 2023.Ac-cording to the occurrence of GDM,the patients were divided into incidence group(30 cases)and non-incidence group(120 cases).Lo-gistic regression model was used to analyze the independent risk factors of occurrence of GDM,and the obtained independent risk factors were included in Lasso-Nomogram prediction model.The efficacy of Lasso-Nomogram prediction model was evaluated by C index,receiver opera-tor characteristic(ROC)curve,and calibration curve in R software.Results Logistic regression analysis showed that aged or more than 35 years old,body mass index(BMI)24 kg/m2 or above before pregnancy,preference for sweets,family history of diabetes,and polycystic ovary syndrome(PCOS)before pregnancy were all independent risk factors for GDM(OR=12.091,59.000,38.000,37.667,12.346,all P＜0.05).ROC curve showed that the predictive value of aged or more than 35 years old,BMI 24 kg/m2 or above before pregnancy,pref-erence for sweets,family history of diabetes,and PCOS before pregnancy for GDM was high.Lasso-Nomogram prediction model was estab-lished based on the influencing factors,and C index of calibration curve was 0.830,which showed that Lasso-Nomogram prediction model had good discrimination,and the values of area under ROC curve in modeling group and verification group were 0.825 and 0.923,respec-tively,which showed that Lasso-Nomogram model had good prediction energy efficiency.Conclusion Lasso-Nomogram prediction model based on the independent influencing factors of GDM can directly predict the incidence probability of GDM.
Gestational diabetes mellitusInfluencing factorLasso-Nomogram prediction model
NETL
NSTL
万方数据
