妇女产后1年体重滞留列线图预测模型的构建与评估

Construction and evaluation of weight retention nomogram prediction model for women 1 year after delivery

目的分析妇女产后1年体重滞留的影响因素,构建并评估产后1年体重滞留的列线图预测模型。方法于2010年9月—2011年2月在广州市越秀区和白云区妇幼保健院招募468名孕晚期孕妇,追踪至产后1年,采用自制问卷收集孕妇的基本人口学信息,3天24小时膳食回顾法调查孕晚期膳食摄入,测量妇女分娩前及产后1年体重。根据产后1年滞留体重是否>0 kg,将研究对象分为产后1年体重滞留组和体重恢复组,运用Logistic回归筛选产后1年体重滞留的影响因素。应用R 4.2.3软件构建列线图预测模型,并采用受试者工作特征曲线、校准曲线及Hosmer-Lemeshow拟合优度检验、临床决策曲线评估模型的区分度、准确度及临床适用性。结果模型训练集329名研究对象产后1年体重滞留率为68.09%,滞留体重的四分位数水平为5.0(3.0,10.0)kg。经Logistic回归分析,以孕前体质指数、孕期增重、产次、孕次、产后0~6月喂养方式构建列线图预测模型,模型区分度(AUC_(训练)=0.778,AUC_(验证)=0.767)和准确度(Hosmer-Lemeshow检验:P训练=0.946,P_(验证)=0.891)较好。结论基于妇女孕前体质指数、孕期增重、产次、孕次、产后0~6月喂养方式构建的产后1年体重滞留列线图模型具有良好的区分度、准确度和临床适用度。

OBJECTIVE To analyze the influencing factors of body weight retention in woman at 1 year postpartum,and to construct and evaluate a nomogram prediction model for postpartum 1-year weight retention.METHODS From September 2010 to February 2011,468 pregnant women in the third trimester were recruited from Yuexiu District and Baiyun District Maternal and Child Health Hospital in Guangzhou,and followed up to 1 year postpartum.The basic demographic information of pregnant women was collected by self-made questionnaire.Dietary intake in the third trimester was investigated by 3-day 24-hour dietary review.The weight of women before delivery and one year after delivery were measured.According to whether the weight retention at 1 year postpartum is greater than 0 kg,the study subjects were divided into the 1-year postpartum weight retention group and weight recovery group.Logistic regression analysis were used to screen the influencing factors of weight retention at 1 year postpartum.R 4.2.3 software was used to construct the nomogram prediction model.The subject working characteristic curve,calibration curve,Hosmer-Lemeshow goodness of fit test and clinical decision curve were used to evaluate the model's differentiation,accuracy and clinical applicability.RESULTS Among 329 subjects in the model training set,the 1-year postpartum weight retention was 68.09%,and the median and quartile levels of retained body weight were 5.0(3.0,10.0)kg.After Logistic analysis,a nomogram prediction model was constructed based on five factors:pre-pregnancy body mass index(BMI),pregnancy weight gain,parity,gravitity,0-6 months postpartum feeding pattern.The model had good discrimination(AUC_(training)=0.778,AUC_(testing)=0.767)and accuracy(Hosmer-Lemeshow test:P_(training)=0.946,P_(testing)=0.891).CONCLUSION The 1-year postnatal weight retention nomogram model based on women's pre-pregnancy BMI,pregnancy weight gain,parity,gravitity,0-6 months postpartum feeding pattern has good differentiation,accuracy and clinical applicability.