To construct and validate machine learning-based prediction models for the risk of excessive gestational weight gain(EGWG),and to compare the predictive performance of different models, and select the optimal one.

Using a convenience sampling method,a total of 623 puerpera who delivered at term and received prenatal check-ups at a tertiary grade A hospital in Beijing from December 2024 to June 2025 were selected as the study subjects.Univariate and multivariate Logistic regression were used to screen predictive factors.Prediction models were constructed by using six machine learning algorithms:Logistic regression,support vector machine,gradient boosting machine,neural network,extreme gradient boosting,and categorical boosting.Model performance was evaluated by using metrics,such as the area under the receiver operating characteristic curve(AUC),F1 score,accuracy,sensitivity,specificity,precision,and Brier Score to determine the optimal model,and a web calculator was developed.The SHapley additive explanations(SHAP) method was employed to interpret the contribution of each variable to the outcome in the best model.

Among 623 puerpera,198 (31.8%) cases experienced excessive gestational weight gain. Multivariate analysis results showed that pre-pregnancy body mass index,sedentary time during pregnancy,gestational diabetes mellitus,advanced maternal age,depressive symptoms during pregnancy,insufficient physical activity during pregnancy,and primiparity were risk factors for excessive gestational weight gain.Among the six machine learning models,the gradient boosting machine model demonstrated the best predictive performance(training set: AUC=0.836,F1=0.669;test set:AUC=0.756,F1=0.609).Calculation of SHAP values for the gradient boosting machine model revealed that pre-pregnancy body mass index,sedentary time during pregnancy,and gestational diabetes mellitus were the three key feature variables for predicting excessive gestational weight gain.

This study constructed a machine learning algorithm-based risk prediction model and web calculator for excessive gestational weight gain,providing an effective tool for early clinical identification and prevention.This contributes to the implementation of targeted clinical interventions and improves the effectiveness of gestational weight management.

To analyze the status quo of sleep quality of pregnant women in late pregnancy,and to explore the correlation between sleep quality in late pregnancy and pregnancy outcomes.

Pregnant women who visited the Obstetrics Clinic of Nantong Maternal and Child Health Hospital from November 2023 to March 2024 were selected as the research subjects.Pittsburgh Sleep Quality Index(PSQI),Epworth Sleepiness Scale(ESS),and Snail Sleep APP were used to evaluate the sleep quality of pregnant women at 28-32 weeks and 32-36 weeks,respectively.Clinical data related to pregnancy outcomes were obtained from the hospital's electronic medical record system.Spearman rank correlation analysis was used for univariate correlation analysis.

A total of 219 pregnant women were included.The correlation analysis results showed that the sleep quality of pregnant women in the late pregnancy showed a downward trend with the increase of gestational weeks[the awakening time after falling asleep was longer(P<0.001),and the total PSQI score was significantly increased(P<0.001)].The sleep efficiency of pregnant women in late pregnancy was negatively correlated with the risk of macrosomia(r=-0.17,P=0.011).Pregnant women with high frequency of nocturia had an increased risk of adverse outcomes such as postpartum hemorrhage(r=0.12,P=0.046),premature birth(r=0.14,P=0.043),neonatal asphyxia(r=0.25,P<0.001),and admission to the neonatal intensive care unit(r=0.13,P=0.045).

The sleep quality of pregnant women gradually decreases in the late pregnancy.The sleep quality of pregnant women directly affects the health status of the mother and offspring.

To explore subjective cognitive attitudes toward lifestyle changes among patients after metabolic and bariatric surgery and inform tailored interventions.

Based on the Integrated Theory of Health Behavior Change,Q methodology was applied.A 28-item Q-sample was developed via literature review and semi-structured interviews.Sixteen patients within 1-12 months post-surgery completed Q-sorting.Factor extraction and analysis were conducted.

Fourteen valid Q-sorts were analyzed.Three distinct attitude types were identified,accounting for 62.0% of total variance:proactive,passive acceptance,and cognitive conflict.These types differed in health cognition,behavioral motivation,and contextual adaptability.

Patients after bariatric metabolic surgery exhibit heterogeneous attitudes toward lifestyle changes.Identifying cognitive typologies can inform stratified nursing education and behavioral support strategies to enhance long-term adherence and rehabilitation outcomes.

To construct a Nomogram prediction model for intraoperative acquired pressure injury in elderly patients with brain tumors based on limb muscle strength.

A total of 536 elderly patients with brain tumors who underwent surgical treatment in our hospital from January 2022 to June 2025 were enrolled and randomly divided into a training set(n=366) and a validation set(n=170) at a 7∶3 ratio.Based on the occurrence of intraoperative acquired pressure injury,patients in the training set were further classified into an injury group and a non-injury group.Clinical data and limb muscle strength data were collected.Univariate analysis was performed.The least absolute shrinkage and selection operator(LASSO) regression algorithm was used to identify characteristic variables associated with intraoperative acquired pressure injury.Multivariate Logistic regression analysis was conducted to determine independent predictors of intraoperative acquired pressure injury.A Nomogram prediction model was then constructed based on these independent predictors.The calibration of the model was evaluated by using the Hosmer-Lemeshow test and calibration curves.Clinical utility was assessed through decision curve analysis.Predictive performance was evaluated by plotting receiver operating characteristic curves and calculating the area under the curve.External validation was performed by using the validation set.

Logistic regression analysis results showed that diabetes,operative time,intraoperative temperature variation,albumin level,C-reactive protein level,and limb muscle strength grade were independent predictive factors of intraoperative acquired pressure injury in elderly patients with brain tumors.The calibration curves for both the training and validation sets showed good fit with the ideal curve.The Hosmer-Lemeshow test indicated good consistency between predicted and observed probabilities,demonstrating satisfactory model accuracy(training set:χ²=6.697,P=0.570;validation set:χ²=5.464,P=0.707).Decision curve analysis revealed that within a threshold probability range of 0 to 1,the Nomogram model provided a higher net clinical benefit compared to any single predictor in both the training and validation sets,indicating strong clinical utility.The Area under receiver operating characteristic curves for the training set was 0.883(95%CI 0.845-0.920) and for the validation set was 0.879(95%CI 0.820-0.937),indicating excellent predictive.

The Nomogram prediction model for intraoperative acquired pressure injury in elderly patients with brain tumors,incorporating diabetes,operative time,intraoperative temperature variation,albumin level,C-reactive protein level,and limb muscle strength grade,demonstrates good calibration,clinical utility,and predictive performance.The Nomogram prediction model of intraoperative acquired pressure injury in elderly patients with brain tumors based on diabetes,operation time,intraoperative temperature variation,albumin level,C-reactive protein,and limb muscle strength has good calibration,clinical practicality,and prediction efficiency.This model holds significant clinical value for identifying high-risk patients and guiding preventive nursing interventions.

To identify growth trajectory of weight by using latent class growth modeling,and to analyze the risk of secondary PICC catheter displacement in very low birth weight(VLBW) infants with different trajectories.

By a convenience sampling method,very low birth weight infants who underwent PICC placement admitted to Shanxi Bethune Hospital from January 2022 to December 2024 were selected as the survey subjects.Latent class growth modeling was used to identify the weight growth trajectory of the infants after catheterization for 4 weeks.Multivariate Cox regression analysis was used to identify the risk factors for secondary catheter displacement in the infants.The survival curves were drawn to compare the occurrence of secondary catheter displacement in infants with different weight growth trajectories.

A total of 212 very low birth weight infants were included.According to their weight growth trajectory after catheterization,they were divided into three types:decline-then-rise type(38.2%),rise-then-decline type(17.5%),and relatively stable type(44.3%).Multivariate Cox regression analysis results showed that decline-then-rise type[HR=4.290,95%CI (1.853,9.923)],and rise-then-decline type[HR=3.399,95%CI(1.352,8.535)]were risk factors for secondary catheter displacement in very low birth weight infants.The survival curve results showed that with the relatively stable type as a reference,the risk of catheter displacement was higher in the decline-then-rise type(χ²=6.786,P=0.005) and rise-then-decline type(χ²=9.861,P<0.05).There was no statistically significant difference in terms of the risk of catheter displacement between the decline-then-rise type and the rise-then-decline type(χ²=0.507,P=0.476).

The weight growth trajectory of very low birth weight infants after catheterization included three types:decline-then-rise type,rise-then-decline type,and relatively stable type.Decline-then-rise type and rise-then-decline type are risk factors for secondary catheter displacement in very low birth weight infants.Risk stratification of PICC catheter secondary displacement based on different weight growth trajectories could provide a basis for early intervention by medical staff.

To explore the relationship between interoceptive sensitivity and impaired fasting glucose(IFG) in early pregnancy.

A cross-sectional study method was conducted.A total of 233 pregnant women who established their prenatal records in the first trimester in Wuxi Maternal and Child Health Care Hospital were selected as the research subjects.Their demographic,clinical,obstetric,and nutritional data were collected.Interoceptive sensitivity was assessed by using the Multidimensional Assessment of Interoceptive Awareness,version 2(MAIA-2).Participants were divided into an impaired fasting glucose group and a normal blood glucose group based on their fasting blood glucose levels.Multivariate Logistic regression analysis was used to analyze the association between interoceptive sensitivity in early pregnancy and impaired fasting glucose.

The incidence of impaired fasting glucose in 233 pregnant women was 18.0%(42 cases).After adjusting for relevant confounding factors,the multivariable Logistic regression analysis results showed that interoceptive sensitivity was an independent protective factor against impaired fasting glucose in early pregnancy(OR=0.520,95%CI 0.275-0.984,P=0.044).

Higher interoceptive sensitivity in early pregnancy has a protective effect against impaired fasting glucose.This finding suggests that enhancing interoceptive sensitivity in pregnant women may be a potential new target for the prevention and management of abnormal fasting glucose in early pregnancy.