Postpartum anemia risk prediction model and use method thereof
By constructing a risk prediction model for postpartum anemia, using latent category trajectories and Logistic regression models to identify key factors, the problem of difficult prediction of postpartum anemia risks in the existing technology is solved, the prediction accuracy and efficiency are improved, and targeted intervention measures are provided.
Patent Information
- Application Number
- CN202411120968.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-12-03
AI Technical Summary
The lack of effective risk prediction and intervention measures for postpartum anemia in the prior art, which makes it difficult to predict and control the risk of postpartum anemia in early stage.
Construct a risk prediction model for postpartum anemia risk prediction, collect relevant basic factors and influencing indicators, use latent category trajectory model and Logistic regression model to identify the change trajectory of vitamins and minerals, screen out key factors, and build multiple risk prediction models for postpartum anemia to improve prediction Accuracy.
It improves the accuracy and efficiency of risk prediction of postpartum anemia, and provides targeted guidelines to help pregnant women take preventive measures in advance to reduce the risk of postpartum anemia.
Smart Images

Figure CN119069126A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of biomedical technology, and in particular to a postpartum anemia risk prediction model and a method for using the model. Background Art
[0002] Some potential threats and risks that may cause adverse effects can be predicted in advance, such as the risk of postpartum anemia, so that some preventive measures can be taken in advance to avoid adverse effects. Existing studies have shown that the increased risk of postpartum anemia may be related to factors such as the pregnant woman's pre-pregnancy BMI, age, parity, use of supplements during pregnancy, and anemia during pregnancy. [1-3] However, there is still a lack of targeted guidelines and reliable risk screening tools for postpartum anemia, and how to effectively predict and intervene in the occurrence of postpartum anemia at an early stage remains to be explored. Summary of the invention
[0003] Based on this, it is necessary to provide a postpartum anemia risk prediction model and its use method to address the above technical issues.
[0004] In a first aspect, the present application provides a method for constructing a postpartum anemia risk prediction model. The method comprises:
[0005] (I) Collect the basic factors related to the occurrence of postpartum anemia;
[0006] (II) Select multiple influencing indicators that are involved in or affect the risk of postpartum anemia: (a) vitamin content in the blood, (b) mineral content in the blood;
[0007] (III) selecting N healthy pregnant women with singleton fetuses in the second trimester as subjects, using a latent class trajectory model to identify the change trajectory of each of the multiple influencing indicators of each subject during the risk prediction period, and clustering the subjects based on the change trajectory to obtain multiple subject categories for each of the influencing indicators;
[0008] (IV) using a Logistic regression model to analyze the first dependency relationship between the risk of postpartum anemia in the risk prediction period and the change trajectory of each of the multiple influencing indicators, and screening out candidate influencing indicators with a high dependency relationship with postpartum anemia;
[0009] (V) dividing the risk prediction period into a plurality of impact time periods, and using a logistic regression model to analyze a second dependency relationship between the risk of postpartum anemia and each candidate impact indicator in each impact time period;
[0010] (VI) Based on the second dependency relationship, the key factors that affect the risk of postpartum anemia are screened out from the multiple candidate influencing indicators, and multiple postpartum anemia risk prediction models are constructed based on the basic factors and the key factors.
[0011] In one embodiment, the method has one or more features selected from the following:
[0012] (1) the plurality of influencing indicators include vitamin A, vitamin E, vitamin B1, vitamin B2, vitamin B3, vitamin B5, vitamin B6, vitamin B9, vitamin D, iron, magnesium, selenium, copper and zinc;
[0013] (2) The risk prediction period is from the second trimester to 48 hours after delivery;
[0014] (3) The multiple affected time periods include mid-pregnancy, late pregnancy and postpartum; the mid-pregnancy is from 14 to 27 weeks of pregnancy; the late pregnancy is from 28 weeks of pregnancy to delivery; the postpartum is a preset time period after delivery;
[0015] In one embodiment, the method has one or more features selected from the following:
[0016] (1) In the process of identifying the change trajectories of the multiple influencing indicators of each subject in the risk prediction period by using the latent class trajectory model, a coordinate system of time and the influencing indicators is constructed, wherein time is used as the abscissa and the influencing indicators are used as the ordinate, and the change trajectories include "first decrease and then increase", "first increase and then decrease", "stable change", "stable change-low level", "stable change-medium level", "decrease", "increase" and "low level", etc.;
[0017] (2) The logistic analysis of the first dependency relationship between the content change trajectory of a certain influencing index and the risk of postpartum anemia takes the "stable change" or "stable change-low level" or "low level" change trajectory group as the reference group, and the other change trajectories of the influencing index as the experimental group. If there is a significant difference in the risk of postpartum anemia between the experimental group and the reference group, then the influencing index is the influencing index with a high dependency relationship with postpartum anemia.
[0018] (3) The basic factors that affect the risk of postpartum anemia include age, pre-pregnancy BMI, parity, anemia during pregnancy, use of folic acid supplements, and use of iron supplements; the key factors that affect the risk of postpartum anemia include vitamin B1 (mid-pregnancy), vitamin B3 (mid-pregnancy), vitamin B6 (mid-pregnancy), iron (late pregnancy), and copper (late pregnancy).
[0019] As used herein, the term "vitamin B1 (mid-pregnancy)" is defined as the content of vitamin B1 in the blood of a subject in mid-pregnancy; the term "vitamin B3 (mid-pregnancy)" is defined as the content of vitamin B3 in the blood of a subject in mid-pregnancy; the term "vitamin B6 (mid-pregnancy)" is defined as the content of vitamin B6 in the blood of a subject in mid-pregnancy; the term "iron (third trimester)" is defined as the content of iron in the blood of a subject in third trimester; the term "copper (third trimester)" is defined as the content of copper and iron in the blood of a subject in third trimester.
[0020] In a second aspect, the present application provides a postpartum anemia risk prediction model, which is constructed by the method described in any one of the first aspects.
[0021] In one embodiment, the postpartum anemia risk prediction model includes multiple models, and the multiple models specifically include a prediction model, a basic model, a multivariate model and a univariate model; the postpartum anemia risk prediction model is used to predict the postpartum anemia risk of the subject based on the basic factors and key factors of the subject.
[0022] In one embodiment, the model influencing index of the basic model includes a basic factor, which is expressed by the following formula:
[0023]
[0024] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use.
[0025] In one embodiment, the univariate model includes a vitamin B1 univariate model, a vitamin B3 univariate model, a vitamin B6 univariate model, an iron univariate model, a copper univariate model, and the like.
[0026] In one embodiment, the model impact index of the vitamin B1 univariate model includes vitamin B1, which is expressed by the following formula:
[0027]
[0028] Among them, x1 is vitamin B1 (mid-pregnancy).
[0029] In one embodiment, the model impact index of the vitamin B3 univariate model includes vitamin B3, which is expressed by the following formula:
[0030]
[0031] Among them, x1 is vitamin B3 (mid-pregnancy).
[0032] In one embodiment, the model impact index of the vitamin B6 univariate model includes vitamin B6, which is expressed by the following formula:
[0033]
[0034] Among them, x1 is vitamin B6 (mid-pregnancy).
[0035] In one embodiment, the model impact index of the iron univariate model includes iron, which is expressed by the following formula:
[0036]
[0037] Among them, x1 is iron (late pregnancy).
[0038] In one embodiment, the model influencing index of the copper univariate model includes copper, which is expressed by the following formula:
[0039]
[0040] Among them, x1 is copper (late pregnancy).
[0041] In one embodiment, the multivariate model includes a vitamin B1 model, a vitamin B3 model, a vitamin B6 model, an iron model, a copper model, etc.
[0042] In one embodiment, the model influencing index of the vitamin B1 model includes basic factors and vitamin B1, which is expressed by the following formula:
[0043]
[0044] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B1 (mid-pregnancy).
[0045] In one embodiment, the model influencing index of the vitamin B3 model includes basic factors and vitamin B3, which is expressed by the following formula:
[0046]
[0047] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy).
[0048] In one embodiment, the model influencing index of the vitamin B6 model includes basic factors and vitamin B6, which is expressed by the following formula:
[0049]
[0050] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B6 (mid-pregnancy).
[0051] In one embodiment, the model influencing index of the iron model includes basic factors and iron, which is expressed by the following formula:
[0052]
[0053] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is iron (late pregnancy).
[0054] In one embodiment, the model influencing index of the copper model includes basic factors and copper, which is expressed by the following formula:
[0055]
[0056] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is copper (late pregnancy).
[0057] In one embodiment, the multivariate model includes a first model, a second model, a third model, a fourth model, a fifth model, etc.
[0058] In one embodiment, the model influencing index of the first model includes basic factors, vitamin B1, vitamin B3, iron and copper, which are expressed by the following formula:
[0059]
[0060] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B1 (mid-pregnancy); x7 is vitamin B3 (mid-pregnancy); x8 is iron (late pregnancy); x9 is copper (late pregnancy).
[0061] In one embodiment, the model influencing index of the second model includes basic factors, vitamin B3, iron and copper, which are expressed by the following formula:
[0062]
[0063] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); x7 is iron (late pregnancy); x8 is copper (late pregnancy).
[0064] In one embodiment, the model influencing index of the third model includes basic factors, vitamin B1, vitamin B3, vitamin B6 and iron, which are expressed by the following formula:
[0065]
[0066] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); x7 is vitamin B1 (mid-pregnancy); x8 is vitamin B6 (mid-pregnancy); x9 is iron (late pregnancy).
[0067] In one embodiment, the model influencing index of the fourth model includes basic factors, vitamin B3, vitamin B6 and iron, which are expressed by the following formula:
[0068]
[0069] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); x7 is vitamin B6 (mid-pregnancy); x8 is iron (late pregnancy).
[0070] In one embodiment, the model influencing index of the fifth model includes basic factors, vitamin B3 and iron, which are expressed by the following formula:
[0071]
[0072] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); x7 is iron (late pregnancy).
[0073] In one embodiment, the model influencing index of the prediction model includes basic factors and key factors, which are expressed by the following formula:
[0074]
[0075] Among them, x1 is age; x2 is BMI before pregnancy; x3 is parity; x4 is the use of folic acid supplements; x5 is the use of iron supplements; x6 is anemia during pregnancy; x7 is vitamin B3 (mid-pregnancy); x8 is vitamin B1 (mid-pregnancy); x9 is vitamin B6 (mid-pregnancy); x 10 Iron (late pregnancy); x 11 For copper (late pregnancy).
[0076] In a third aspect, the present application provides a method for evaluating the risk of postpartum anemia in a subject, wherein the method comprises: using the postpartum anemia risk prediction model described in the second aspect to obtain a probability value of postpartum anemia in the subject;
[0077] In one embodiment, the greater the probability value of postpartum anemia occurring in the subject, the higher the risk of postpartum anemia occurring.
[0078] In a fourth aspect, the present application provides a system for executing the method described in the third aspect, the system comprising:
[0079] (1) a storage component for storing data, wherein the storage component has instructions for determining an assessment of a subject stored therein;
[0080] (2) a computer processor for processing data, wherein the computer processor is coupled to the storage component and configured to execute instructions stored in the storage component to receive the subject data and analyze the subject data;
[0081] (3) A display component for displaying evaluation information of the subject.
[0082] In one embodiment, the system is a diagnostic system for performing a computer-implemented method for assessing a subject's risk of developing postpartum anemia;
[0083] Wherein, the computer executes the postpartum anemia risk prediction model described in the second aspect.
[0084] In one embodiment, the computer executes the following steps:
[0085] a) receiving input subject data, wherein the data comprises basic factors and key factors of the subject obtained from the method described in the first aspect;
[0086] b) analyzing and calculating the probability value of the occurrence of postpartum anemia in the subject based on the method described in the first aspect;
[0087] Optionally, the greater the probability value of postpartum anemia occurring in the subject, the higher the risk of postpartum anemia occurring.
[0088] In a fifth aspect, the present application provides a kit comprising multiple influencing indicators for determining involvement in or influencing the risk of postpartum anemia.
[0089] In one embodiment, the kit is used to assess the risk of developing postpartum anemia in a subject.
[0090] In one embodiment, the multiple influencing indicators involved in or affecting the risk of postpartum anemia include: vitamin B1 (mid-pregnancy), vitamin B3 (mid-pregnancy), vitamin B6 (mid-pregnancy), iron (late pregnancy) and copper (late pregnancy).
[0091] In a sixth aspect, the present application provides a composition for use in the preparation of a medicine or health food for preventing or treating postpartum anemia, characterized in that the composition comprises vitamin B3 and iron.
[0092] In one embodiment, the composition comprises 12 mg of vitamin B3 and 24 mg of iron.
[0093] In one embodiment, the composition comprises 12 mg of vitamin B3 and 29 mg of iron.
[0094] In one embodiment, the composition further comprises vitamin B1, vitamin B6 and copper.
[0095] In one embodiment, the composition includes 1.4 mg of vitamin B1, 12 mg of vitamin B3, 2.2 mg of vitamin B6, 24 mg of iron and 0.9 mg of copper.
[0096] In one embodiment, the composition includes 1.5 mg of vitamin B1, 12 mg of vitamin B3, 2.2 mg of vitamin B6, 29 mg of iron and 0.9 mg of copper.
[0097] In one embodiment, the composition further comprises vitamin B2, vitamin B5, vitamin B9, vitamin B 12 , vitamin A, vitamin D, vitamin E, vitamin C, calcium, magnesium, iodine, zinc, and / or selenium.
[0098] In one embodiment, the composition comprises 1.2 mg of vitamin B1, 1.2 mg of vitamin B2, 12 mg of vitamin B3, 6.0 mg of vitamin B5, 2.2 mg of vitamin B6, 600 μg of vitamin B9, 12 2.9μg, vitamin A 660μg, vitamin D 10μg, vitamin E 14mg, vitamin C 100mg, calcium 800mg, magnesium 370mg, iron 18mg, iodine 230μg, zinc 10.5mg, selenium 65μg, copper 0.91mg.
[0099] In one embodiment, the composition comprises 1.4 mg vitamin B1, 1.3 mg vitamin B2, 12 mg vitamin B3, 6.0 mg vitamin B5, 2.2 mg vitamin B6, 600 μg vitamin B9, 122.9μg, vitamin A 730μg, vitamin D 10μg, vitamin E 14mg, vitamin C 115mg, calcium 800mg, magnesium 370mg, iron 25mg, iodine 230μg, zinc 10.5mg, selenium 65μg, copper 0.9mg.
[0100] In one embodiment, the composition comprises 1.4 mg vitamin B1, 1.5 mg vitamin B2, 12 mg vitamin B3, 6.0 mg vitamin B5, 2.2 mg vitamin B6, 600 μg vitamin B9, 12 2.9μg, vitamin A 730μg, vitamin D 10μg, vitamin E 14mg, vitamin C 115mg, calcium 800mg, magnesium 370mg, iron 29mg, iodine 230μg, zinc 10.5mg, selenium 65μg, copper 0.9mg.
[0101] In a seventh aspect, the present application provides a medicine or a health food, characterized in that the medicine or the health food comprises vitamin B3 and iron.
[0102] In one embodiment, the medicine or health food includes 12 mg of vitamin B3 and 24 mg of iron.
[0103] In one embodiment, the medicine or health food includes 12 mg of vitamin B3 and 29 mg of iron.
[0104] In one embodiment, the medicine or health food further comprises vitamin B1, vitamin B6 and copper.
[0105] In one embodiment, the medicine or health food includes 1.4 mg of vitamin B1, 12 mg of vitamin B3, 2.2 mg of vitamin B6, 24 mg of iron and 0.9 mg of copper.
[0106] In one embodiment, the medicine or health food includes 1.5 mg of vitamin B1, 12 mg of vitamin B3, 2.2 mg of vitamin B6, 29 mg of iron and 0.9 mg of copper.
[0107] In one embodiment, the composition further comprises vitamin B2, vitamin B5, vitamin B9, vitamin B 12 , vitamin A, vitamin D, vitamin E, vitamin C, calcium, magnesium, iodine, zinc, and / or selenium.
[0108] In one embodiment, the composition comprises 1.2 mg of vitamin B1, 1.2 mg of vitamin B2, 12 mg of vitamin B3, 6.0 mg of vitamin B5, 2.2 mg of vitamin B6, 600 μg of vitamin B9, 12 2.9μg, vitamin A 660μg, vitamin D 10μg, vitamin E 14mg, vitamin C 100mg, calcium 800mg, magnesium 370mg, iron 18mg, iodine 230μg, zinc 10.5mg, selenium 65μg, copper 0.91mg.
[0109] In one embodiment, the composition comprises 1.4 mg vitamin B1, 1.3 mg vitamin B2, 12 mg vitamin B3, 6.0 mg vitamin B5, 2.2 mg vitamin B6, 600 μg vitamin B9, 12 2.9μg, vitamin A 730μg, vitamin D 10μg, vitamin E 14mg, vitamin C 115mg, calcium 800mg, magnesium 370mg, iron 25mg, iodine 230μg, zinc 10.5mg, selenium 65μg, copper 0.9mg.
[0110] In one embodiment, the composition comprises 1.4 mg vitamin B1, 1.5 mg vitamin B2, 12 mg vitamin B3, 6.0 mg vitamin B5, 2.2 mg vitamin B6, 600 μg vitamin B9, 12 2.9μg, vitamin A 730μg, vitamin D 10μg, vitamin E 14mg, vitamin C 115mg, calcium 800mg, magnesium 370mg, iron 29mg, iodine 230μg, zinc 10.5mg, selenium 65μg, copper 0.9mg.
[0111] In an eighth aspect, the present application provides a method for preventing or treating postpartum anemia, characterized in that a composition comprising vitamin B3 and iron is administered to a pregnant woman.
[0112] In one embodiment, the method comprises administering 12 mg NE / d of vitamin B3 and 24 mg / d of iron to pregnant women at 14-27 weeks of gestation; and / or the method comprises administering 12 mg NE / d of vitamin B3 and 29 mg / d of iron to pregnant women at ≥28 weeks of gestation.
[0113] In one embodiment, the composition further comprises vitamin B1, vitamin B6 and copper.
[0114] In one embodiment, the method is to administer vitamin B1 1.4 mg / d, vitamin B3 12 mg NE / d, vitamin B6 2.2 mg / d, iron 24 mg / d and copper 0.9 mg / d to pregnant women at 14-27 weeks of gestation; and / or the method is to administer vitamin B1 1.5 mg / d, vitamin B3 12 mg NE / d, vitamin B6 2.2 mg / d, iron 29 mg / d and copper 0.9 mg / d to pregnant women at ≥28 weeks of gestation.
[0115] In one embodiment, the composition further comprises vitamin B2, vitamin B5, vitamin B9, vitamin B 12 , vitamin A, vitamin D, vitamin E, vitamin C, calcium, magnesium, iodine, zinc, and / or selenium.
[0116] In one embodiment, the method has one or more of the following features:
[0117] (1) Pregnant women ≤ 14 weeks of gestation should receive vitamin B1 1.2 mg / d, vitamin B2 1.2 mg / d, vitamin B3 12 mgNE / d, vitamin B5 6.0 mg / d, vitamin B6 2.2 mg / d, vitamin B9 600 μg DFE / d, and vitamin B 12 2.9μg / d, vitamin A 660μg RAE / d, vitamin D 10μg / d, vitamin E 14mg / d, vitamin C 100mg / d, calcium 800mg / d, magnesium 370mg / d, iron 18mg / d, iodine 230μg / d, zinc 10.5mg / d, selenium 65μg / d, copper 0.91mg / d;
[0118] (2) Pregnant women at 14-27 weeks of gestation should receive 1.4 mg / d of vitamin B1, 1.3 mg / d of vitamin B2, 12 mg / d of vitamin B3, 6.0 mg / d of vitamin B5, 2.2 mg / d of vitamin B6, 600 μg DFE / d of vitamin B9, and 1.4 mg / d of vitamin B1. 12 2.9μg / d, vitamin A 730μg RAE / d, vitamin D 10μg / d, vitamin E 14mg / d, vitamin C 115mg / d, calcium 800mg / d, magnesium 370mg / d, iron 25mg / d, iodine 230μg / d, zinc 10.5mg / d, selenium 65μg / d, copper 0.9mg / d;
[0119] (3) Pregnant women at 28 weeks of gestation should receive vitamin B1 1.4 mg / d, vitamin B2 1.5 mg / d, vitamin B3 12 mgNE / d, vitamin B5 6.0 mg / d, vitamin B6 2.2 mg / d, vitamin B9 600 μg DFE / d, and vitamin B 12 2.9μg / d, vitamin A 730μg RAE / d, vitamin D 10μg / d, vitamin E 14mg / d, vitamin C 115mg / d, calcium 800mg / d, magnesium 370mg / d, iron 29mg / d, iodine 230μg / d, zinc 10.5mg / d, selenium 65μg / d, copper 0.9mg / d.
[0120] In the above-mentioned postpartum anemia risk prediction model and its use method, basic factors related to the occurrence of postpartum anemia can be collected through literature research, and the trajectory of changes in vitamin and mineral content in blood during pregnancy can be identified through the latent class trajectory model; influencing indicators related to the risk of postpartum anemia are included to construct a logistic regression model; the key trajectory of vitamins and minerals related to the risk of postpartum anemia is determined through logistic regression analysis, and the key time points are further determined. Multiple screenings can be performed to obtain multiple key factors that meet the requirements for the degree of correlation with the risk of postpartum anemia, which can improve the accuracy of the construction of the postpartum anemia risk prediction model. Further, the risk of postpartum anemia can be predicted based on the constructed postpartum anemia risk prediction model, which can improve the accuracy, efficiency and simplicity of postpartum anemia risk prediction. BRIEF DESCRIPTION OF THE DRAWINGS
[0121] Figure 1 A schematic diagram of the changing trajectory of each influencing indicator in the risk prediction cycle provided in the embodiment of the present application;
[0122] Figure 2 A schematic diagram of the AUC comparison of multiple models of the postpartum anemia risk prediction model provided in the embodiments of the present application. DETAILED DESCRIPTION
[0123] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application. In an exemplary embodiment, a method for constructing a postpartum anemia risk prediction model is provided. In this embodiment, the method comprises the following steps:
[0124] (I) Literature review to collect basic factors associated with the occurrence of postpartum anemia; these basic factors include age, pre-pregnancy BMI, parity, anemia during pregnancy, folic acid supplementation, and iron supplements.
[0125] The keywords "pregnancy" or "postpartum" and "anemia" and "risk factors" were combined to search for relevant studies on factors affecting anemia during pregnancy and childbirth in literature databases (e.g., CNKI, Wanfang, VIP, Pubmed, Web of science, and Ovid, etc.). The P value in the results of univariate analysis or multivariate regression model analysis in the literature was used as the inclusion criterion. If the P value was ≤ 0.05, the factor was considered to have an impact on anemia during pregnancy and childbirth and was selected as the basic factor. Studies conducted in Xinjiang, Shaanxi, Jiangsu and other regions have shown that [4-6][4-7] , maternal age (P = 0.027), pre-pregnancy BMI (P = 0.004), and parity (P < 0.01) are important influencing factors for anemia during pregnancy, so the above factors were included as the basic factors of the model. Anemia is a nutritional disease and is related to insufficient intake of nutrients such as iron and folic acid. Studies have shown that iron and folic acid supplementation during pregnancy can reduce the risk of anemia (P < 0.01) [8] Therefore, the use of iron and folic acid supplements was included in the model as basic factors. Through the above process, age, pre-pregnancy BMI, parity, iron supplementation and folic acid supplementation were finally included in the model as basic factors.
[0126] (II) Select multiple influencing indicators that are involved in or affect the risk of postpartum anemia: (a) vitamin content in the blood, (b) mineral content in the blood.
[0127] Among them, the multiple influencing indicators may include the vitamin content in the blood and the mineral content in the blood, and specifically may include vitamin A, vitamin E, vitamin B1, vitamin B2, vitamin B3, vitamin B5, vitamin B6, vitamin B9, vitamin D, iron, magnesium, selenium, copper and zinc, etc. At this stage, the multiple influencing indicators can be regarded as multiple initial screening influencing indicators.
[0128] (III) N healthy pregnant women with singleton pregnancy in the second trimester were selected as subjects, and the latent class trajectory model was used to identify the change trajectory of each of the multiple influencing indicators of each subject during the risk prediction period, and the subjects were clustered based on the change trajectory of each of the multiple influencing indicators to obtain multiple subject categories for each influencing indicator.
[0129] First, multiple subjects can be selected, and the subjects can be healthy pregnant women with a single fetus in the second trimester (14 to 27 weeks of pregnancy), where N is a positive integer, such as 200, 400 and 600, etc., and the latent class trajectory model analysis includes N valid samples, and some human components that have a certain correlation with postpartum anemia can be selected as influencing indicators for predicting the probability of postpartum anemia, such as the content of vitamin A, vitamin E, vitamin B1, vitamin B2, vitamin B3, vitamin B5, vitamin B6, vitamin B9, vitamin D, iron, magnesium, selenium, copper and zinc in the blood of the subjects. Next, the latent class trajectory model can be used to identify the change trajectory of each influencing indicator (primary screening influencing indicator) in the risk prediction period according to the content change information of the multiple influencing indicators (primary screening influencing indicators) in the blood of the subjects during the risk prediction period. There can be multiple indicator change types, that is, there are multiple change trajectories for each primary screening influencing indicator during the risk prediction period. Based on the change trajectory of each primary screening influencing indicator during the risk prediction period, multiple subjects can be clustered and classified, and subjects with consistent change trajectories can be divided into one group. Specifically, the risk prediction period can be from the second trimester to 48 hours after delivery. The latent class trajectory model can be used to identify the change trajectory of the vitamin and mineral concentration in the subject's blood from the second trimester to 48 hours after delivery. The subjects are clustered according to the change characteristics of the vitamin and mineral content in the subjects from the second trimester to 48 hours after delivery, and people with similar content change characteristics are divided into the same category, and the categories are defined according to the characteristics of the trajectory change. Figure 1 As shown in the figure, the change trajectories (i.e., change characteristics) of vitamin A, vitamin E, vitamin B1, vitamin B2 and iron levels in the subjects were classified into three categories, which were defined as "first decrease and then increase", "first increase and then decrease" and "stable change" according to their change characteristics; the change trajectories of vitamin B3 and magnesium levels were divided into three categories, which were defined as "decrease", "first increase and then decrease" and "stable change" respectively; the change trajectories of vitamin B6 and selenium levels were divided into three categories, which were defined as "decrease", "increase" and "stable change" respectively; the change trajectories of vitamin B5 and vitamin B9 levels were divided into four categories, which were defined as "first decrease and then increase", "first increase and then decrease", "stable change-low level" and "stable change-medium level" according to their change characteristics; the change trajectories of copper and vitamin D levels were divided into two categories, which were defined as "low level" and "high level" according to the change trajectory characteristics of copper content, and were defined as "first decrease and then increase" and "stable change" according to the change trajectory characteristics of vitamin D content; the change trajectory of zinc content was divided into five categories, which were defined as "decrease", "increase", "first increase and then decrease", "stable change-low level" and "stable change-medium level" respectively.
[0130] (IV) The Logistic regression model was used to analyze the first dependency relationship between the risk of postpartum anemia in the risk prediction period and the change trajectory of each of the multiple influencing indicators, and the candidate influencing indicators with a high dependency relationship with the occurrence of postpartum anemia were screened out.
[0131] Among them, for each influencing indicator, the first dependency relationship between the probability of postpartum anemia in the subject (i.e., risk value) and the change trajectory of the influencing indicator under multiple change trajectories can be obtained by the Logistic regression model, and the first dependency relationship can be used to represent the change of the probability of postpartum anemia in the subject with the change of each influencing indicator. The logistic analysis of the first dependency relationship between the content change trajectory of a certain influencing indicator and the risk of postpartum anemia is based on the "stable change" or "stable change-low level" or "low level" change trajectory group as the reference group, and the other change trajectories of the influencing indicator are the experimental groups. If there is a significant difference in the risk of postpartum anemia between the experimental group and the reference group, then the influencing indicator is an influencing indicator with a high dependency relationship with postpartum anemia.
[0132] On this basis, a first preset dependency threshold can be preset, and the first preset dependency threshold can be the minimum dependency requirement for selecting multiple candidate influencing indicators from multiple influencing indicators. Further, the influencing indicator corresponding to the first dependency relationship exceeding the first preset dependency threshold can be determined as the candidate influencing indicator. The first dependency relationship corresponding to each influencing indicator can be referred to in Table 1. For example, vitamin B1, vitamin B 3、 Taking vitamin B6, iron and copper as examples, the risk of postpartum anemia in the vitamin B1 "first decrease and then increase" group was significantly lower than that in the "stable change" group (OR: 0.246, 95% CI: 0.102-0.595, P: 0.002), while no statistically significant results were found in the "first increase and then decrease" group. The risk of postpartum anemia in the vitamin B3 "decline" group was lower than that in the "stable change" group (OR: 0.215, 95% CI: 0.078-0.591, P: 0.003), while no statistically significant results were found in the "first increase and then decrease" group. The risk of postpartum anemia in the vitamin B6 "decline" group was significantly lower than that in the "stable change" group (OR: 0.255, 95% CI: 0.070-0.934, P: 0.039), while no statistically significant results were found in the "increase" group. The risk of postpartum anemia in the iron "first increase and then decrease" group was significantly lower than that in the "stable change" group (OR: 0.195, 95% CI: 0.101-0.377, P: 0.000), while no statistically significant results were found in the "first decrease and then increase" group. The risk of postpartum anemia in the copper "high level" group was significantly increased compared with the "low level" group (OR: 2.415, 95% CI: 1.216-4.797, P: 0.012).
[0133] Table 1 The first dependency relationship corresponding to the influencing indicators
[0134]
[0135]
[0136]
[0137] Among them, model 1 a It is a univariate model, in which the independent variables only include the corresponding impact indicators; Model 2 b It is a regression model that takes basic factors into consideration, i.e., a multivariate model. The independent variables in the multivariate model include basic factors and corresponding influencing indicators. The basic factors may include but are not limited to age, pre-pregnancy BMI, educational level, parity, iron supplementation, folic acid supplementation, mode of delivery, and iron.
[0138] (V) The risk prediction cycle is divided into multiple impact time periods, and the Logistic regression model is used to analyze the secondary dependency relationship between the risk of postpartum anemia and each impact indicator in each impact time period.
[0139] Among them, the risk prediction period can be from the second trimester to 48 hours after delivery, and the risk prediction period of postpartum anemia can be divided into three influencing time periods: the second trimester, the third trimester and the postpartum period; when multiple candidate influencing indicators are screened out from multiple influencing indicators, the change of the index value of each candidate influencing indicator in different influencing time periods can be different, so the index value of each candidate influencing indicator in different influencing time periods can be further obtained. In one possible implementation, in each influencing time period, the influencing time period can be set to at least one data collection cycle, and the index value of each candidate influencing indicator in each data collection cycle is obtained, and further, the average value of multiple index values corresponding to multiple data collection cycles is determined as the index value of each candidate influencing indicator in the influencing time period; in another possible implementation, the index value of the median time point of each influencing time period can be obtained as the index value of each candidate influencing indicator in each influencing time period. This application does not limit the method for obtaining the index value of each candidate influencing indicator in different influencing time periods.
[0140] Among them, for each candidate influencing indicator, the second dependency relationship between the probability of postpartum anemia (i.e., risk value) of the subject and the index value of the candidate influencing indicator in different influencing time periods can be obtained through the Logistic regression model; on this basis, a second preset dependency threshold can be set in advance, and the second preset dependency threshold can be the minimum dependency requirement for predicting the risk of postpartum anemia. In general, the test level in the logistic regression analysis can be set in advance to 0.05, and the P value of the analysis result of each candidate influencing indicator obtained can be compared with the test level. When the P value of the influencing indicator in a certain influencing time period and the logistic regression analysis result of the risk of postpartum anemia shows ≤0.05, it can be considered that the difference is statistically significant, that is, the influencing indicator in the influencing time period is one of the key factors affecting the probability of postpartum anemia. The second dependency relationship information between the probability of postpartum anemia (i.e., risk value) of the subject and the index value of the candidate influencing indicator in different influencing time periods can be found in Table 2, with vitamin B1, vitamin B 3、 Taking vitamin B6, iron and copper as examples, with the increase of vitamin B1 (OR: 0.810, 95% CI: 0.668-0.982, P: 0.032), vitamin B3 (OR: 0.849, 95% CI: 0.729-0.989, P: 0.036) and vitamin B6 (OR: 0.889, 95% CI: 0.802-0.985, P: 0.024) levels in the second trimester, the risk of postpartum anemia decreased. The iron content in mid-pregnancy (OR: 0.996, 95% CI: 0.992-0.999, P: 0.020), late pregnancy (OR: 0.995, 95% CI: 0.991-0.999, P: 0.014) and postpartum (OR: 0.986, 95% CI: 0.980-0.991, P: 0.000) was negatively correlated with the risk of postpartum anemia, that is, as the iron content in the mother's body increased, the risk of postpartum anemia decreased. The copper content in the mother's body during late pregnancy (OR: 4.530, 95% CI: 1.391-14.750, P: 0.012) and after delivery (OR: 3.370, 95% CI: 1.031-11.020, P: 0.044) was positively correlated with the risk of postpartum anemia, that is, as the copper content in the body increased, the risk of postpartum anemia increased.
[0141] Table 2 Second dependency relationships corresponding to candidate impact indicators
[0142]
[0143]
[0144]
[0145] Among them, model 1 a It is a univariate model, in which the independent variables only include the corresponding impact indicators; Model 2 b It is a multivariate model that takes into account basic factors. The independent variables in the multivariate model include basic factors and corresponding influencing indicators. The basic factors may include but are not limited to age, pre-pregnancy BMI, educational level, parity, iron supplementation, folic acid supplementation, mode of delivery, iron, etc.
[0146] (VI) Based on the second dependency, the key factors that affect the risk of postpartum anemia are screened out from multiple candidate influencing indicators, and multiple postpartum anemia risk prediction models are constructed based on basic factors and key factors.
[0147] Among them, the postpartum anemia risk prediction model can predict the probability of postpartum anemia. The determined key factors can be the target impact index within the target impact time period. In a possible implementation, the postpartum anemia risk prediction model of the target impact index can be constructed based on the probability of postpartum anemia in the subject and the dependency of the basic factors and key factors within the target impact time period. The independent variables included in the postpartum anemia risk prediction model are basic factors and key factors. For postpartum anemia, the basic factors may include but are not limited to age, pre-pregnancy BMI, education, parity, iron supplementation, folic acid supplementation, mode of delivery, iron, etc. The key factors include but are not limited to vitamin B3 (mid-pregnancy), vitamin B1 (mid-pregnancy), vitamin B6 (mid-pregnancy), iron (late pregnancy) and copper (late pregnancy), etc.
[0148] The indicators (basic factors + key factors) determined according to the above steps are used as independent variables, and the subject's postpartum anemia condition is used as the dependent variable to construct a logistic regression model equation as a postpartum anemia risk prediction model. When used, the actual conditions of the indicators in the subject's model are input into the postpartum anemia risk prediction model, and the probability of the subject developing postpartum anemia can be obtained as a risk prediction value.
[0149] In an exemplary embodiment, the method for constructing a postpartum anemia risk prediction model in the embodiment of the present application meets the following characteristics:
[0150] (1) Multiple impact indicators include vitamin A, vitamin E, vitamin B1, vitamin B2, vitamin B5, vitamin B9, vitamin D, iron, magnesium, selenium, copper and zinc;
[0151] (2) The risk prediction period is from the second trimester to 48 hours after delivery;
[0152] (3) Multiple affected time periods include mid-pregnancy, late pregnancy, and postpartum; mid-pregnancy is from 14 to 27 weeks of pregnancy; late pregnancy is from 28 weeks of pregnancy to delivery; postpartum is the preset time period after delivery;
[0153] In an exemplary embodiment, the method for constructing a postpartum anemia risk prediction model in the embodiment of the present application meets the following characteristics:
[0154] (1) Using the latent class trajectory model to identify the change trajectories of multiple influencing indicators of each subject during the risk prediction period, a coordinate system of time and influencing indicators is constructed, wherein time is used as the horizontal axis and the influencing indicator is used as the vertical axis. The change trajectories include "first decrease and then increase", "first increase and then decrease", "stable change", "stable change-low level", "stable change-medium level", "decrease", "increase" and "low level", etc.;
[0155] (2) The logistic analysis of the first dependency relationship between the content change trajectory of a certain influencing indicator and the risk of postpartum anemia takes the "stable change" or "stable change-low level" or "low level" change trajectory group as the reference group, and the other change trajectories of the influencing indicator as the experimental group. If there is a significant difference in the risk of postpartum anemia between the experimental group and the reference group, then the influencing indicator is an influencing indicator with a high dependency relationship with postpartum anemia.
[0156] (3) Basic factors that affect the risk of postpartum anemia include age, pre-pregnancy BMI, parity, anemia during pregnancy, use of folic acid supplements, and use of iron supplements; key factors that affect the risk of postpartum anemia include vitamin B3 (mid-pregnancy), vitamin B1 (mid-pregnancy), vitamin B6 (mid-pregnancy), iron (late pregnancy), and copper (late pregnancy).
[0157] In the method of this embodiment, the trajectory of changes in vitamin and mineral content during pregnancy can be identified through the latent class trajectory model; the basic factors related to the risk of postpartum anemia are included to construct a logistic regression model; the key trajectory of vitamins and minerals related to the risk of postpartum anemia is determined through logistic regression analysis, and the key time points are further determined. Multiple screenings can be performed to obtain multiple key factors that meet the requirements for the degree of correlation with the risk of postpartum anemia, which can improve the accuracy of the construction of the postpartum anemia risk prediction model. Further, the risk of postpartum anemia can be predicted based on the constructed postpartum anemia risk prediction model, which can improve the accuracy, efficiency and simplicity of postpartum anemia risk prediction.
[0158] In an exemplary embodiment, the construction of a postpartum anemia risk prediction model may include:
[0159] The indicators (basic factors + key factors) determined according to the above steps are used as independent variables, and the subject's postpartum anemia condition is used as the dependent variable to construct a logistic regression model equation as a postpartum anemia risk prediction model. When in use, the actual conditions of the indicators in the subject's model are input into the postpartum anemia risk prediction model, and the probability of the subject developing postpartum anemia can be obtained as a risk prediction value. The postpartum anemia risk prediction model can include: the postpartum anemia risk prediction model includes multiple models, and the multiple models specifically include prediction models, basic models, multivariate models and univariate models. The postpartum anemia risk prediction model is used to predict the risk of postpartum anemia in the subject based on the changes in the subject's influencing indicators.
[0160] The model influencing index of the basic model includes basic factors, which are expressed by the following formula:
[0161]
[0162] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use.
[0163] The univariate models include vitamin B1 univariate model, vitamin B3 univariate model, vitamin B6 univariate model, iron univariate model, copper univariate model and the like.
[0164] The model influence index of the vitamin B1 univariate model includes vitamin B1 and is expressed by the following formula:
[0165]
[0166] Among them, x1 is vitamin B1 (mid-pregnancy).
[0167] The model influence indicator of the vitamin B3 univariate model includes vitamin B3 and is expressed by the following formula:
[0168]
[0169] Among them, x1 is vitamin B3 (mid-pregnancy).
[0170] The model influence indicator of the vitamin B6 univariate model includes vitamin B6 and is expressed by the following formula:
[0171]
[0172] Among them, x1 is vitamin B6 (mid-pregnancy).
[0173] The model-influencing indicator of the iron univariate model includes iron and is expressed as follows:
[0174]
[0175] Among them, x1 is iron (late pregnancy).
[0176] The model-influencing indicator of the copper univariate model includes copper and is expressed as follows:
[0177]
[0178] Among them, x1 is copper (late pregnancy).
[0179] Multivariate models include vitamin B1 model, vitamin B3 model, vitamin B6 model, iron model, copper model, etc.
[0180] The model influencing index of the vitamin B1 model includes basic factors and vitamin B1, which is expressed by the following formula:
[0181]
[0182] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B1 (mid-pregnancy).
[0183] The model influencing index of the vitamin B3 model includes basic factors and vitamin B3, which is expressed by the following formula:
[0184]
[0185] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy).
[0186] The model influencing index of the vitamin B6 model includes basic factors and vitamin B6, which is expressed by the following formula:
[0187]
[0188] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B6 (mid-pregnancy).
[0189] The model influencing index of the iron model includes basic factors and iron, which is expressed by the following formula:
[0190]
[0191] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is iron (late pregnancy).
[0192] The model influencing indicators of the copper model include basic factors and copper, which are expressed by the following formula:
[0193]
[0194] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is copper (late pregnancy).
[0195] The multivariate model includes a first model, a second model, a third model, a fourth model, a fifth model, and the like.
[0196] Among them, the model influencing indicators of the first model include basic factors, vitamin B1, vitamin B3, iron and copper, which are expressed by the following formula:
[0197]
[0198] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B1 (mid-pregnancy); x7 is vitamin B3 (mid-pregnancy); x8 is iron (late pregnancy); x9 is copper (late pregnancy).
[0199] The model influencing indicators of the second model include basic factors, vitamin B3, iron and copper, which are expressed by the following formula:
[0200]
[0201] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); x7 is iron (late pregnancy); x8 is copper (late pregnancy).
[0202] The model influencing indicators of the third model include basic factors, vitamin B1, vitamin B3, vitamin B6 and iron, which are expressed by the following formula:
[0203]
[0204] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); x7 is vitamin B1 (mid-pregnancy); x8 is vitamin B6 (mid-pregnancy); x9 is iron (late pregnancy).
[0205] The model influencing indicators of the fourth model include basic factors, vitamin B3, vitamin B6 and iron, which are expressed by the following formula:
[0206]
[0207] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); x7 is vitamin B6 (mid-pregnancy); x8 is iron (late pregnancy).
[0208] The model influencing indicators of the fifth model include basic factors, vitamin B3 and iron, which are expressed by the following formula:
[0209]
[0210] Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); x7 is iron (late pregnancy).
[0211] The model influencing indicators of the prediction model include basic factors and key factors, which are expressed by the following formula:
[0212]
[0213] Among them, x1 is age; x2 is BMI before pregnancy; x3 is parity; x4 is the use of folic acid supplements; x5 is the use of iron supplements; x6 is anemia during pregnancy; x7 is vitamin B3 (mid-pregnancy); x8 is vitamin B1 (mid-pregnancy); x9 is vitamin B6 (mid-pregnancy); x 10 Iron (late pregnancy); x 11 For copper (late pregnancy).
[0214] The model has one or more of the following characteristics:
[0215] (1) After vitamin B1 (mid-pregnancy), vitamin B3 (mid-pregnancy), vitamin B6 (mid-pregnancy), iron (late pregnancy) and copper (late pregnancy) were included in the basic model, the AUC of the model was improved.
[0216] (2) The prediction model had the best performance, and the AUC of the prediction model was significantly higher than that of the basic model (AUC difference: 0.154, P: 0.0003).
[0217] The postpartum anemia risk prediction model can be shown in Table 3, and the prediction performance parameters of the postpartum anemia risk prediction model can be shown in Table 4. The AUC comparison of multiple models of the postpartum anemia risk prediction model can be shown in Table 4. Figure 2 As shown, in the case where the multiple target influence indicators can be all the target influence indicators of the postpartum anemia risk, the postpartum anemia risk prediction model of the postpartum anemia risk can be called a prediction model; the univariate model of the target influence indicator can be named after the current influence indicator. Figure 2It can include model AUC comparison between the basic model and the univariate model, model AUC comparison between the basic model and the prediction model, model AUC comparison between the prediction model and the univariate model, and model AUC comparison between different multivariate postpartum anemia risk prediction models.
[0218] Table 3 Postpartum anemia risk prediction model
[0219]
[0220] Table 4 Prediction performance parameters of the postpartum anemia risk prediction model
[0221] Model AUC 95% CI P Sensitivity (%) Specificity (%) Basic Model 0.607 0.536~0.676 0.0075 53.33 71.58 <![CDATA[Vitamin B1 Model]]> 0.623 0.552~0.691 0.0018 40.95 81.05 <![CDATA[Vitamin B3 Model]]> 0.645 0.574~0.711 0.0002 64.76 60.00 <![CDATA[Vitamin B6 Model]]> 0.617 0.546~0.685 0.0034 61.90 63.16 Iron Model 0.662 0.592~0.727 <0.0001 72.38 57.89 Copper Model 0.640 0.569~0.706 0.0004 53.33 75.79 <![CDATA[Vitamin B1 (second trimester of pregnancy)]]> 0.570 0.498~0.639 0.0874 86.67 28.42 <![CDATA[Vitamin B3 (second trimester of pregnancy)]]> 0.596 0.524~0.664 0.0189 75.24 48.42 <![CDATA[Vitamin B6 (second trimester of pregnancy)]]> 0.585 0.513~0.654 0.0354 65.71 52.63 Iron (late pregnancy) 0.647 0.576~0.713 0.0002 61.90 68.42 Copper (late pregnancy) 0.590 0.519~0.659 0.0257 53.33 66.32 Model 5 0.746 0.680~0.805 <0.0001 67.86 80.00 Model 4 0.741 0.674~0.800 <0.0001 80.00 64.21 Model 3 0.710 0.642~0.772 <0.0001 55.24 82.11 Model 2 0.700 0.632~0.763 <0.0001 55.24 78.95 Model 1 0.697 0.628~0.760 <0.0001 59.05 73.68 Prediction Model 0.762 0.696~0.819 <0.0001 72.38 70.53
[0222] Sensitivity, specificity and AUC are indicators for evaluating the performance of a binary classification model, and are used to measure the predictive ability and discrimination of the model. Among them, sensitivity reflects the ability of the prediction model to find patients, and specificity reflects the ability of the prediction model to determine non-patients. When both sensitivity and specificity are required to be high, it is necessary to use the ROC curve to find the balance point of sensitivity and specificity. The ROC curve is a curve drawn based on a series of different binary classification cutoff values, with sensitivity as the ordinate and 1-specificity as the abscissa. AUC is the area under the ROC curve. The closer the AUC is to 1.0, the higher the authenticity and accuracy of the prediction model. Therefore, the evaluation of the model prediction performance mainly refers to the numerical value of AUC, and the improvement of the prediction performance is judged according to the size of AUC. The method of this embodiment can be based on at least one of the multiple target influencing indicators that meet the requirements of the degree of association with the risk of postpartum anemia, and the multiple postpartum anemia risk prediction models can be constructed, which can adapt to various situations and improve the accuracy, efficiency and universality of the prediction of postpartum anemia risk.
[0223] In an exemplary embodiment, the present application embodiment proposes a method for evaluating the risk of postpartum anemia in a subject, which may include: using the multiple postpartum anemia risk prediction models constructed above to obtain the probability value of postpartum anemia in the subject. In a possible implementation, the greater the probability value of postpartum anemia in the subject, the higher the risk of postpartum anemia in the subject; the smaller the probability value of postpartum anemia in the subject, the lower the risk of postpartum anemia in the subject.
[0224] Among them, the multiple postpartum anemia risk prediction models may include prediction models, basic models, multivariate models and univariate models. For different postpartum anemia risk prediction models, the corresponding influencing indicators are different. Taking the prediction model in the multiple postpartum anemia risk prediction models as an example, the index values of the basic factors and key factors of the subject can be obtained, and the basic factors may include but are not limited to age, pre-pregnancy BMI, parity, iron supplementation, folic acid supplementation, etc. The key factors may include anemia during pregnancy, vitamin B1 (mid-pregnancy), vitamin B3 (mid-pregnancy), vitamin B6 (mid-pregnancy), iron (late pregnancy) and copper (late pregnancy); Next, the index values of the basic factors and key factors of the subject can be input into the prediction model, and the probability value of the occurrence of postpartum anemia in the subject can be obtained by using the prediction model. According to the probability value, the risk of postpartum anemia in the subject can be judged. The greater the probability value of postpartum anemia in the subject, the higher the risk of postpartum anemia in the subject; the smaller the probability value of postpartum anemia in the subject, the lower the risk of postpartum anemia in the subject. The working processes of the other models in the multiple postpartum anemia risk prediction models except the prediction model can refer to the prediction model.
[0225] In an exemplary embodiment, the embodiment of the present application proposes a postpartum anemia risk prediction system, which can be used to perform the above-mentioned method for evaluating the risk of postpartum anemia in a subject, that is, it can be used to predict the risk of postpartum anemia in a subject and obtain the probability value of postpartum anemia in the subject. The postpartum anemia risk prediction system may include a storage component, a computer processor, and a display component. Among them, the storage component can be used to store data, and the stored data may include instructions for determining the evaluation of the subject stored therein; the computer processor can be used to process data, and the computer processor is coupled to the storage component and is configured to execute instructions stored in the storage component to accept subject data and analyze subject data; the display component can be used to display the subject's evaluation information.
[0226] In an exemplary embodiment, the postpartum anemia risk prediction system is a diagnostic system for executing a computer-implemented method for assessing the risk of postpartum anemia in a subject. Wherein, the computer executes the above-mentioned postpartum anemia risk prediction model. In a possible implementation, the computer can execute the method for assessing the risk of postpartum anemia in a subject provided in the above-mentioned embodiment of the present application, which may specifically include: a) receiving the input subject data, the data containing the basic factors and key factors of the subject obtained from the method for constructing a postpartum anemia risk prediction model provided in the above-mentioned embodiment of the present application; b) analyzing and calculating the probability value of the occurrence of postpartum anemia in the subject based on the method for constructing a postpartum anemia risk prediction model provided in the above-mentioned embodiment of the present application. Wherein, taking the prediction model in the multiple postpartum anemia risk prediction models as an example, the index values of the basic factors and key factors of the subject can be obtained, and the basic factors may include but are not limited to age, pre-pregnancy BMI, parity, iron supplementation, folic acid supplementation, etc. The key factors may include anemia during pregnancy, vitamin B1 (mid-pregnancy), vitamin B3 (mid-pregnancy), vitamin B6 (mid-pregnancy), iron (late pregnancy) and copper (late pregnancy); next, the index values of the subject's basic factors and key factors can be input into the prediction model, and the prediction model can be used to obtain the probability value of the subject's postpartum anemia. According to the probability value, the risk of the subject developing postpartum anemia can be judged. The greater the probability value of the subject's postpartum anemia, the higher the risk of the subject's postpartum anemia; the smaller the probability value of the subject's postpartum anemia, the lower the risk of the subject's postpartum anemia. The action process of other models except the prediction model in the multiple postpartum anemia risk prediction models can be referred to the prediction model.
[0227] In an exemplary embodiment, the present application embodiment provides a kit, which contains multiple influencing indicators for determining participation in or affecting the risk of postpartum anemia. The kit is used to assess the risk of postpartum anemia in a subject. In a possible implementation, the multiple influencing indicators may include: vitamin B1 (mid-pregnancy), vitamin B3 (mid-pregnancy), vitamin B6 (mid-pregnancy), iron (late pregnancy) and copper (late pregnancy).
[0228] In an exemplary embodiment, the present application embodiment studies the intervention effect of a vitamin and mineral composition on the occurrence of postpartum anemia, and provides a use of the composition in the preparation of a medicine or health food for preventing or treating postpartum anemia. The specific research method is as follows:
[0229] (1) Trial population:
[0230] Included in the scope:
[0231] Convenient sampling was used to select healthy pregnant women with singleton pregnancy in the second trimester (14 weeks to 27+6 weeks of gestation) who received prenatal checkups at a maternal and child health care institution in Guangzhou from April to June 2023. According to the results of prenatal checkups during pregnancy, Hb ≥ 110 g / L was included in the group (no anemia during pregnancy was diagnosed).
[0232] Exclusion criteria:
[0233] ① Patients with serious diseases such as heart, cerebrovascular, liver, kidney, digestive tract and mental illness
[0234] ② Those with allergic constitution or allergic to the test sample
[0235] ③ Taking items related to the function being tested in a short period of time may affect the judgement of the results
[0236] ④ Those who fail to take the test sample according to the standard or whose data are incomplete and affect the judgment of efficacy or safety
[0237] (2) Experimental design, i.e. grouping:
[0238] This project included 45 subjects, using an intergroup control design, with 15 people in each group. The intervention started from the second trimester of pregnancy to before delivery (about 6 months). During the trial, the original eating habits were not changed and the diet was normal. The specific groups are as follows:
[0239] Control group: received routine dietary guidance during pregnancy
[0240] Experimental Group 1:
[0241] From the second trimester to before delivery, take a multinutrient supplement tablet once a day (14 to 27 weeks of pregnancy: vitamin B3 12 mg NE / d, iron 24 mg / d; ≥ 28 weeks of pregnancy: vitamin B3 12 mg NE / d, iron 29 mg / d)
[0242] Experimental Group 2:
[0243] From the second trimester to before delivery, take 2 multinutrient supplement tablets once a day (14 to 27 weeks of pregnancy: vitamin B1 1.4 mg / d, vitamin B3 12 mg NE / d, vitamin B6 2.2 mg / d, iron 24 mg / d, copper 0.9 mg / d; ≥ 28 weeks of pregnancy: vitamin B1 1.5 mg / d, vitamin B3 12 mg NE / d, vitamin B6 2.2 mg / d, iron 29 mg / d, copper 0.9 mg / d)
[0244] (3) Observation indicators and efficacy evaluation:
[0245] ① During the trial, ask and record the general condition of the pregnant woman (mental state, sleep, diet, defecation, urination, blood pressure, etc.) and symptoms of adverse reactions;
[0246] ② Collect blood samples from mothers 24 hours after delivery to test hemoglobin levels;
[0247] ③Hemoglobin <110g / L was diagnosed as postpartum anemia. The incidence of postpartum anemia was compared among the three groups.
[0248] (4) Test results:
[0249] Table 5 Gestational age and hemoglobin level of pregnant women at enrollment
[0250]
[0251] As shown in Table 5, there were no significant differences in gestational age and hemoglobin levels between the control group and experimental groups 1 and 2 at the time of enrollment, and the hemoglobin levels were all >110 g / L, and none of the subjects was diagnosed with anemia at the time of enrollment.
[0252] In addition, the mental state, sleep, diet, bowel movements, and blood pressure of the pregnant women in the experimental and control groups were normal from the second trimester of pregnancy until delivery, and no obvious adverse reactions such as nausea, vomiting, abnormal liver and kidney function occurred.
[0253] Table 6 Comparison of postpartum hemoglobin levels between the experimental group and the control group
[0254] Grouping Hb(g / L) F P Control group (n=15) 111.00±8.06 Experimental group 1 (n=15) 118.00±5.39 8.782 0.001 Experimental group 2 (n=15) 120.19±4.89
[0255] Table 7 Comparison of the incidence of postpartum anemia between the experimental group and the control group
[0256] Grouping Number of people with postpartum anemia Postpartum anemia incidence (%) <![CDATA[χ 2 ]]> P Control group (n=15) 7 46.67 Experimental group 1 (n=15) 2 13.33 7.021 0.049 Experimental group 2 (n=15) 1 6.67
[0257] Among them, it can be seen from Tables 6 and 7 that the incidence of postpartum anemia in test group 1 is lower than that in the control group, and the incidence of postpartum anemia in test group 2 is lower than that in test group 1, indicating that the composition has preventive and therapeutic effects on postpartum anemia, and the composition of test group 2 is better than the composition of test group 1 in preventing or treating postpartum anemia.
[0258] Although the specific embodiments of the present invention have been described in detail, it will be understood by those skilled in the art that various modifications and changes may be made to the details according to all the teachings that have been published, and these changes are within the scope of protection of the present invention. The entire invention is given by the attached claims and any equivalents thereof.
[0259] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.
[0260] References
[0261] [1]ABEBE GT,KURE MA,YADETA TA,et al.Immediate postpartum anemia and associated factors among women admitted to maternity ward at publichospitals in HarariRegional State,Eastern Ethiopia:A facility-based cross-sectional study[J].Frontiers inglobal women's health,2022,3:916245.
[0262] [2]AGMASSIE GA,ALAMNEH GD,AYICHEH MW,et al.The magnitude and associated factors of immediate postpartum anemia among women who gave birthin east Gojjam zonehospitals,northwest-Ethiopia,2020[J].PloS one,2023,18(3):e0282819.
[0263] [3]ZEWDU D, TANTU T, ALI R, et al. Prevalence and predictors of postpartum anemia aftercaesarean delivery in Ethiopia: A retrospective analysis of risk factors[J]. Midwifery, 2023,123:103707.
[0265] [4] Cao Yingying. Prevalence and influencing factors of anemia in pregnant women in Taicang City from 2012 to 2016[D], 2018.
[0266] [5] Sun Yijie, Wang Xiaoli. Analysis of the prevalence and influencing factors of anemia during pregnancy in Xinjiang in 2019 (in Chinese) Maternal and Child Health Care[J]. 2023, 38(17): 3304-8.
[0267] [6] Zhang Birong, Fan Yahui, Liu Sijiao, et al. Analysis of the current status and influencing factors of anemia in pregnant women in some areas of Shaanxi Province (in Chinese) Maternal and Child Health Research [J]. 2023, 34(09): 82-8.
[0268] [7] Yu Chunzhi, Zhao Xianli, Kang Ying. High-risk factors for anemia during pregnancy and its impact on pregnancy outcomes (in Chinese) Maternal and Child Health Care [J]. 2017, 32(23): 5827-30.
[0269] [8]DA SILVA LOPES K,YAMAJI N,RAHMAN MO,et al.Nutrition-specific interventions for preventing and controlling anaemia throughout the lifecycle:an overview of systematicreviews[J].The Cochrane database of systematicreviews,2021,9(9):Cd013092.
Claims
1. A method for constructing a postpartum anemia risk prediction model, the method comprising: (I) Collect the basic factors related to the occurrence of postpartum anemia; (II) Select multiple influencing indicators that are involved in or affect the risk of postpartum anemia: (a) vitamin content in the blood, (b) mineral content in the blood; (III) selecting N healthy pregnant women with singleton fetuses in the second trimester as subjects, using a latent class trajectory model to identify the change trajectory of each of the multiple influencing indicators of each subject during the risk prediction period, and clustering the subjects based on the change trajectory to obtain multiple subject categories for each of the influencing indicators; (IV) using a Logistic regression model to analyze the first dependency relationship between the risk of postpartum anemia in the risk prediction period and the change trajectory of each of the multiple influencing indicators, and screening out candidate influencing indicators with a high dependency relationship with postpartum anemia; (V) dividing the risk prediction period into a plurality of impact time periods, and using a logistic regression model to analyze a second dependency relationship between the risk of postpartum anemia and each candidate impact indicator in each impact time period; (VI) Based on the second dependency relationship, the key factors that affect the risk of postpartum anemia are screened out from the multiple candidate influencing indicators, and multiple postpartum anemia risk prediction models are constructed based on the basic factors and the key factors.
2. The method of claim 1, wherein: The method has one or more characteristics selected from the following: (1) The multiple influencing indicators include vitamin A, vitamin E, vitamin B1, vitamin B 2、 Vitamin B3, vitamin B5, vitamin B6, vitamin B9, vitamin D, iron, magnesium, selenium, copper and zinc; (2) The risk prediction period is from the second trimester to 48 hours after delivery; (3) The multiple affected time periods include mid-pregnancy, late pregnancy and postpartum; the mid-pregnancy is from 14 to 27 weeks of pregnancy; the late pregnancy is from 28 weeks of pregnancy to delivery; and the postpartum is a preset time period after delivery.
3. The method of claim 1, wherein: The method has one or more characteristics selected from the following: (1) In the process of identifying the change trajectories of the multiple influencing indicators of each subject during the risk prediction period using the latent class trajectory model, a coordinate system of time and the influencing indicators is constructed, wherein time is used as the abscissa and the influencing indicators are used as the ordinate, and the change trajectories include "first decrease and then increase", "first increase and then decrease", "stable change", "stable change-low level", "stable change-medium level", "decrease", "increase" and "low level", etc.; (2) The logistic analysis of the first dependency relationship between the content change trajectory of a certain influencing index and the risk of postpartum anemia takes the "stable change" or "stable change-low level" or "low level" change trajectory group as the reference group, and the other change trajectories of the influencing index as the experimental group. If there is a significant difference in the risk of postpartum anemia between the experimental group and the reference group, then the influencing index is the influencing index with a high dependency relationship with postpartum anemia. (3) The basic factors that affect the risk of postpartum anemia include age, pre-pregnancy BMI, parity, anemia during pregnancy, use of folic acid supplements, and use of iron supplements; the key factors that affect the risk of postpartum anemia include vitamin B1 (mid-pregnancy), vitamin B3 (mid-pregnancy), vitamin B6 (mid-pregnancy), iron (late pregnancy), and copper (late pregnancy).
4. A postpartum anemia risk prediction model, wherein the model is constructed by the method described in any one of claims 1 to 3.
5. The model of claim 4, wherein: The postpartum anemia risk prediction model includes multiple models, and the multiple models specifically include a prediction model, a basic model, a multivariate model and a univariate model; the postpartum anemia risk prediction model is used to predict the postpartum anemia risk of the subject based on the basic factors and key factors of the subject.
6. The model of claim 5, wherein the univariate model comprises a vitamin B1 univariate model, a vitamin B3 univariate model, a vitamin B6 univariate model, an iron univariate model, a copper univariate model, etc.; Preferably, the model influencing index of the vitamin B1 univariate model includes the vitamin B1, which is expressed by the following formula: in, x1 is the vitamin B1 (mid-pregnancy); Preferably, the model influencing index of the vitamin B3 univariate model includes the vitamin B3, which is expressed by the following formula: Wherein, x1 is the vitamin B3 (mid-pregnancy); Preferably, the model influencing index of the vitamin B6 univariate model includes the vitamin B6, which is expressed by the following formula: Wherein, x1 is the vitamin B6 (mid-pregnancy); Preferably, the model influencing index of the iron univariate model includes the iron, which is expressed by the following formula: wherein x1 is the iron (late pregnancy); and / or Preferably, the model influencing index of the copper univariate model includes the copper, which is expressed by the following formula: Wherein, x1 is the copper (late pregnancy).
7. The model of claim 5, wherein: The multivariate model includes a vitamin B1 model, a vitamin B3 model, a vitamin B6 model, an iron model, a copper model, etc.; Preferably, the model influencing index of the vitamin B1 model includes basic factors and vitamin B1, which is expressed by the following formula: Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B1 (mid-pregnancy); Preferably, the model influencing index of the vitamin B3 model includes basic factors and vitamin B3, which is expressed by the following formula: Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); Preferably, the model influencing index of the vitamin B6 model includes basic factors and vitamin B6, which is expressed by the following formula: Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B6 (mid-pregnancy); Preferably, the model influencing index of the iron model includes basic factors and iron, which is expressed by the following formula: Where x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is iron (late pregnancy); and / or Preferably, the model influencing index of the copper model includes basic factors and copper, which is expressed by the following formula: Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is copper (late pregnancy).
8. The model of claim 5, wherein: The multivariate model includes a first model, a second model, a third model, a fourth model, a fifth model, etc.; Preferably, the model influencing index of the first model includes basic factors, vitamin B1, vitamin B3, iron and copper, which are expressed by the following formula: Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B1 (mid-pregnancy); x7 is vitamin B3 (mid-pregnancy); x8 is iron (late pregnancy); x9 is copper (late pregnancy); Preferably, the model influencing index of the second model includes basic factors, vitamin B3, iron and copper, which are expressed by the following formula: Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); x7 is iron (late pregnancy); x8 is copper (late pregnancy); Preferably, the model influencing index of the third model includes basic factors, vitamin B1, vitamin B3, vitamin B6 and iron, which are expressed by the following formula: Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); x7 is vitamin B1 (mid-pregnancy); x8 is vitamin B6 (mid-pregnancy); x9 is iron (late pregnancy); Preferably, the model influencing index of the fourth model includes basic factors, vitamin B3, vitamin B6 and iron, which are expressed by the following formula: Where x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (second trimester); x7 is vitamin B6 (second trimester); x8 is iron (third trimester); and / or Preferably, the model influencing index of the fifth model includes basic factors, vitamin B3 and iron, which are expressed by the following formula: Among them, x1 is age; x2 is pre-pregnancy BMI; x3 is parity; x4 is folic acid supplement use; x5 is iron supplement use; x6 is vitamin B3 (mid-pregnancy); x7 is iron (late pregnancy).
9. The model of claim 5, wherein: The model influencing index of the prediction model includes basic factors and key factors, which are expressed by the following formula: Among them, x1 is age; x2 is BMI before pregnancy; x3 is parity; x4 is the use of folic acid supplements; x5 is the use of iron supplements; x6 is anemia during pregnancy; x7 is vitamin B3 (mid-pregnancy); x8 is vitamin B1 (mid-pregnancy); x9 is vitamin B6 (mid-pregnancy); x 10 Iron (late pregnancy); x 11 For copper (late pregnancy).
10. A system for executing a method for assessing the risk of postpartum anemia in a subject, the system comprising: (1) a storage component for storing data, wherein the storage component has instructions for determining an assessment of a subject stored therein; (2) a computer processor for processing data, wherein the computer processor is coupled to the storage component and configured to execute instructions stored in the storage component to receive the subject data and analyze the subject data; (3) a display component for displaying the subject's assessment information; Preferably, the system is a diagnostic system for performing a computer-implemented method for assessing a subject's risk of developing postpartum anemia; Preferably, the computer executes the postpartum anemia risk prediction model according to any one of claims 4 to 9; Preferably, the computer executes the following steps: a) receiving input subject data, wherein the data comprises basic factors and key factors of the subject obtained from the method according to any one of claims 1 to 3; b) analyzing and calculating the probability of occurrence of postpartum anemia in a subject according to the method described in any one of claims 1 to 3; Optionally, the probability value of the occurrence of postpartum anemia in the subject is displayed, and the larger the probability value is, the higher the risk of the subject developing postpartum anemia.
11. A kit comprising a plurality of influence indicators for determining participation in or influence on the risk of postpartum anemia; Preferably, the kit is used to assess the risk of postpartum anemia in a subject; Preferably, the multiple influencing indicators involved in or affecting the risk of postpartum anemia include: Vitamin B1 (mid-pregnancy), vitamin B3 (mid-pregnancy), vitamin B6 (mid-pregnancy), iron (late pregnancy), copper (late pregnancy).
12. Use of a composition in preparing a medicine or health food for preventing or treating postpartum anemia, characterized in that: The composition includes vitamin B3 and iron; Preferably, the composition further comprises vitamin B1, vitamin B6, and copper; Preferably, the composition further comprises vitamin B2, vitamin B5, vitamin B9, vitamin B 12 , vitamin A, vitamin D, vitamin E, vitamin C, calcium, magnesium, iodine, zinc, and / or selenium.
