Risk Factors and Nomogram Prediction Model for Iron Deficiency Anemia in Late Pregnancy
10.13471/j.cnki.j.sun.yat-sen.univ（med.sci）.2025.0113
- VernacularTitle:妊娠晚期缺铁性贫血的危险因素和列线图预测模型
- Author:
Lin YANG
1
;
Yixiong WANG
1
;
Hui PAN
1
;
Yang XU
1
Author Information
1. Department of Obstetrics， Yangzhou Maternal and Child Health Hospital， Yangzhou 225002， China
- Publication Type:Journal Article
- Keywords:
late pregnancy;
iron deficiency anemia;
risk factors;
protective factors;
nomogram;
prediction model
- From:
Journal of Sun Yat-sen University(Medical Sciences)
2025;46(1):116-122
- CountryChina
- Language:Chinese
-
Abstract:
ObjectiveThis study aims to identify potential risk factors， establish a nomogram prediction model and propose intervention strategies for iron deficiency anemia （IDA） in late pregnancy. MethodsQuestionnaire surveys and routine blood tests were performed on pregnant women who received regular antenatal check-ups at Yangzhou Maternal and Child Health Hospital from July 2022 to December 2023. Of the 500 cases enrolled， 482 completed the trial and were divided into IDA and non-IDA groups based on the presence of IDA in the third trimester. Univariate analysis and multivariate logistic regression were employed to identify risk factors for IDA， while R software was used to construct a nomogram prediction model. ResultsThe incidence of IDA in late pregnancy was 19.92% （96/482）. Univariate analysis revealed significant associations between IDA in late pregnancy and factors such as economic independence， preconception body mass index （BMI）， number of pregnancies， and regular iron supplementation during pregnancy （P<0.05）. Multivariate logistic regression showed that economic independence （P=0.031， OR=0.583） and regular iron supplementation during pregnancy （P<0.001， OR=5.337） were protective factors against IDA. Conversely， a low preconception BMI （P=0.021， OR=2.375） and three or more pregnancies （P=0.015， OR=2.253） emerged as significant risk factors. The receiver operating characteristic （ROC） curve analysis demonstrated that the area under the curve （AUC） for the nomogram predicting IDA was 0.84， with an optimal cutoff value of -1.481， a sensitivity of 81.2% and a specificity of 75.1%. ConclusionsThe incidence of IDA in late pregnancy is quite high. A low preconception BMI and three or more pregnancies are risk factors， while economic independence and regular iron supplementation are protective factors. These insights have enabled the construction of a prediction model， which， in turn， could provide guidance for prediction and treatment of IDA in late pregnancy.
