To investigate the role and mechanism of CD4⁺ T-cell-derived IFN-γ in host defense against blood-stage Plasmodium berghei (P. berghei) infection.

A tamoxifen-inducible conditional knockout strategy was performed to generate mice lacking IFN-γ specifically in CD4⁺T cells (Ifng^fl/flCD4-CreER^T2). Then after the mice were infected intravenously with luciferase-expressing P. berghei ANKA (P. berghei ANKA-luciferase) strain, parasite burden, survival, and intra-erythrocytic developmental rhythm were monitored dynamically. Flow cytometry and ELISA were carried out respectively to quantify the splenic myeloid subsets and the titers of P. berghei-specific IgG and IgG2c.

CD4⁺T-cell-specific IFN-γ deficiency significantly reduced peripheral parasitemia on day 5 post-infection (P<0.05), showing no effect on survival or 24-hour intra-erythrocytic cycle of the parasite. Ifng^fl/fl CD4-CreER^T2 mice exhibited obviously decreased P. berghei-specific IgG2c (P<0.01) but increased total IgG (P<0.05). Frequencies and absolute numbers of neutrophils, monocytes, and macrophages remained unchanged.

During the process of P. berghei infection, IFN-γ secreted by CD4⁺T cells may decrease total serum IgG level by promoting class switching to IgG2c, and then enhance parasite burden after blood-stage P. berghei infection.

To investigate the effect and underlying mechanisms of curcumin (CUR) on small intestinal injuries associated with severe traumatic infections after combined treatment of ciprofloxacin (CIP) and indomethacin (IND).

Adult male C57BL/6 mice (8 weeks old) were randomly divided into (10 animals for survival rate observation, 10 for other indicators) control group, model group [hemorrhage+ fracture+Pseudomonas aeruginosa (PA) infection], IND+CIP treatment group, and IND+ CIP+CUR treatment group. The survival rates within 72 h post-modelling were observed across all groups. At 24 h post-modelling, peripheral blood samples, small intestinal tissues and mesenteric lymph nodes (MLN) were harvested. Small intestinal length, hemorrhagic status, and adhesions were also observed grossly in each group. Bacterial load was determined via dilute plating. HE staining was used to observe histopathological alterations in intestinal tissues across groups. ELISA was employed to measure the contents of inflammatory cytokines, IL-1β, IL-6, IL-10 and TNF-α in the serum and small intestinal homogenate of each group. Liver, renal, and cardiac functions were evaluated by measuring key biochemical indicators. Flow cytometry was applied to measure the percentages of IFN-γ, IL-4, TGF-β, and IL-17 positive cells within CD4⁺ T cells in the MLN. Real-time quantitative polymerase chain reaction (qPCR) was performed to detect the mRNA expression of forkhead box protein P3(FOXP3) and retinoic acid receptor-related orphan receptor gamma (RORγ)-t in each group.

Both the IND+CIP treatment group and the IND+CIP+CUR treatment group significantly improved survival rates in mice with severe traumatic infections (P<0.05). CUR addition reduced bacterial load in the mice (P<0.01). The IND+CIP treatment group exhibited small intestinal injury, which worsened over time, manifested by elevated contents of TNF-α (P<0.01), IL-6 (P<0.05), and IL-1β (P<0.05) and decreased IL-10 (P<0.01) in the small intestine tissues, accompanied with pathological damage in the small intestine. The IND+CIP+CUR treatment group had significantly reversed small intestinal injury, reduced inflammatory factor levels (P<0.01), and restored normal small intestinal structure, showing no significant differences in relevant blood biochemical indicators when compared to the IND+CIP treatment group. The expression of FOXP3 was increased and that of RORγ-t was decreasedin CD4⁺ T cells from the IND+CIP+CUR treatment group than the IND+CIP treatment group (both P<0.01).

The combination of CUR with IND and CIP can effectively alleviate severe traumatic infections, while antagonize small intestinal damage observed in IND and CIP therapy, with no significant adverse effects on vital organ functions. This mechanism may be related to the balance between regulatory T (Treg) cells and T helper 17 (Th17) cells.

TTo explore the role of circ-gabra5 in the pathogenesis of Alzheimer’s disease (AD) in mice.

APP/PS1 double transgenic AD mice (6 months old) were selected as the experimental group, and C57BL/6 wild-type (WT) mice were enrolled as the control group. The mice were divided into an AD group and an AD mice with circ-gabra5 knockout (AD+circ-gabra5 knockout) group, with 5 mice in each group. After total RNA was extracted from mouse brain, differential expression of circRNA was detected, and real-time quantitative PCR (RT-qPCR) was used to validate the top 6 circRNA genes with differences. Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analyses were performed. Using hippocampal stereo injection of lentivirus to knock out the circ-gabra5 gene, the learning and memory function of mice was detected with Morris water maze test, and the number of Aβ plaques in mouse brain tissue was observed through immunohistochemistry. Transmission electron microscopy was utilized to observe the mitochondrial damage and synaptic count in mouse neurons. Dual luciferase reporter assay and RNA binding protein immunoprecipitation (RIP) assay were employed to verify the interaction between circ-gabra5 and miR-455-3p. Western blotting was apllied to detect the expression level of AKT protein.

There were 273 differentially expressed genes screened between AD mice and WT mice, including 163 up-regulated and 110 down-regulated genes. Circ-gabra5 had the largest fold difference of 1.39 (P<0.05). GO and KEGG analyses suggested that the functional expression differences between AD mice and WT mice were enriched in thyroid hormone, PI3K/AKT, RNA metabolism and other signaling pathways. Knocking out of circ-gabra5 gene in AD mice resulted in obvious improvement in cognitive impairment in Morris water maze test, decreased Aβ plaques by immunohistochemistry, and attenuated neuronal mitochondrial damage and more cortical synapses in AD mice by transmission electron microscopy (P<0.05). According to the starBase database prediction, circ-gabra5 and miR-455-3p have complementary sequences. Dual luciferase reporter assay and RIP assay confirmed the interaction between circ-gabra5 and miR-455-3p. Database bioinformatics analysis predicted that miR-455-3p and AKT have complementary sequences. qPCR indicated that knocking down the circ-gabra5 gene in primary cultured neurons of AD mice resulted in a decrease in miR-455-3p expression, and subsequent Western blot analysis showed a decrease in AKT protein level (P<0.01).

Circ-gabra5 exacerbates cognitive impairment in AD mice by regulating AKT protein through miR-455-3p, increases Aβ deposition in brain tissue, aggravates neuronal mitochondrial damage, and reduces the number of neural synapses.

To investigate the effect of prenatal estriol (E3) exposure on autism spectrum disorders (ASD) and the role of the Notch signaling pathway in the process.

Untargeted metabolomics was performed on plasma samples from 44 mothers of ASD children and 44 matched controls in the Guangxi Zhuang birth cohort. Then E3 was identified as main differential metabolite during pregnancy. SH-SY5Y cells were subjected and divided into a control group and study groups (10, 30, 50 and 100 μmol/L E3), and proteomics was performed to screen the main pathway involved in the process. CCK-8 assay was employed to detect cell proliferation. Western blotting and RT-qPCR were performed to verify the obtained pathway.

Metabolomics analysis for plasma metabolite levels in pregnant women showed that the E3 level in the case group was significantly down-regulated than that of the control group (P<0.01). KEGG enrichment analysis indicated that the Notch signaling pathway was a significantly enriched pathway (P<0.01). CCK-8 assay suggested that the concentration of 100 μmol/L was identified as the most significant dose for E3 in the proliferation inhibition of SH-SY5Y cells (P<0.01). Both qPCR and Western blotting indicated that 100 μmol/L E3 treatment resulted in obvious down-regulation in Jagged1 (P<0.01), Notch1 (P<0.01), and Hes1 (P<0.05) at mRNA and protein levels in SH-SY5Y cells.

Prenatal E3 exposure may be associated with ASD, and its mechanism may be through inhibiting the Notch signaling pathway, and thereby inhibiting neural cell proliferation.

To construct an Escherichia coli (E. coli)-specific reporter phage, T6::Nluc, and evaluate its performance in the rapid diagnosis of E. coli-induced urinary tract infections.

T6::Nluc was constructed by the CRISPR/Cas9 system combined with homologous recombination technology. Then biological characteristics of T6 and T6::Nluc were systematically compared in terms of optimal multiplicity of infection (MOI), one-step growth curves, and lysis efficiency. Time-luminescence curves were plotted to determine the limit of detection (LOD) and the impact of urine matrix. Luminescence quantification was performed to detect the specificity of the phage after being co-cultured with non-E.coli pathogens. In addition, 104 strains of E. coli isolated from clinical urine samples were collected to analyze the detection coverage of T6::Nluc. Finally, 437 clinical urine samples were obtained for clinical validation.

The successful construction of reporter phage T6::Nluc was validated by Sanger sequencing. Its one-step growth curve and optimal MOI differed from those of T6, and the lysis efficiency remained comparable. T6::Nluc exhibited outstanding diagnostic performance, enabling the detection of ≥10⁴ CFU/mL E. coli within 40 min, with complete resistance to interference from urine matrix and excellent pathogen-specific recognition. Additionally, T6::Nluc successfully detected 51 out of 104 clinical E. coli isolates (49.04%). And the results of the detection in clinical urine samples indicated that the phage had a sensitivity of 48.84% and a specificity of 100.00 in the detection.

Reporter phage T6::Nluc is successfully constructed, and can sensitively and specifically detect E. coli causing urinary tract infections, significantly reducing detection time and enhancing efficiency and accuracy.

To investigate the therapeutic effects of ginsenoside Rb1 on non-alcoholic steatohepatitis (NASH) and its mechanisms through modulating gut microbiota and repairing intestinal barriers in liver inflammation and metabolic disorders.

A model NASH mouse was established by high-fat, high-cholesterol diet (HFHCD) feeding. After modeling, the NASH mice were treated with Rb1 via intraperitoneal injection. An in vitro NASH model was induced with oleic acid+cholesterol in HepG2 cells. Serum biochemistry, histopathological observation (HE, Oil Red O, Masson, Sirius Red staining), and Western blotting (inflammatory and tight junction proteins) were performed. Gut microbiota was analyzed with 16S rRNA gene sequencing. The correlations between key microbiota and metabolic phenotypes were analyzed using Spearman correlation analysis and SHapley Additive exPlanations (SHAP) analysis.

Rb1 significantly reduced serum levels of total cholesterol (TC), triglyceride (TG), alanine transaminase (ALT), aspartate aminotransferase (AST) and low-density lipid-cholesterol (LDL-C) in the NASH mice (P<0.05), decreased lipid accumulation and fibrosis in liver tissues, inhibited TNF-α and IL-1β expression, and increased zonula occludens-1（ZO-1) and Occludin levels (P<0.05). It also boosted beneficial bacteria like Lactobacillus and Bifidobacterium, reduced harmful bacteria such as Desulfovibrio, and enhanced gut microbiota diversity (P<0.05). Correlation analysis showed that beneficial bacteria enriched by Rb1 were positively correlated with the improvements of metabolic indicators and negatively with inflammatory factors and liver injury markers (P<0.05). The machine learning model identified ALT as the key predictor for NASH.

Rb1 can inhibit inflammatory response and improve lipid metabolism by modulating gut microbiota and repairing the intestinal barrier, and thus significantly alleviate NASH progression.

To compare the characteristics of tongue coating flora in healthy people, patients with chronic hepatitis B (CHB) and patients with CHB complicated with non-alcoholic fatty liver disease (NAFLD), and to explore the characteristics of tongue coating flora in patients with CHB complicated with NAFLD and its comorbidity mechanism.

From January 2024 to December 2024, 42 with CHB complicated with NAFLD patients with CHB complicated with NAFLD(CHB complicated with NAFLD group) and 34 patients with simple CHB (simple CHB group) were subjected in our department, and 28 healthy individuals were selected and served as control group. Baseline clinical data were collected from all the participants, and intergroup differences were analyzed. After tongue coating samples were harvested for 16S ribosomal RNA (16S rRNA) sequencing using the PacBio Sequel sequencing platform. After sequence quality control, denoising, and amplicon sequence variant (ASV) clustering, the microbial composition, α diversity, and β diversity were analyzed in the 3 sample groups. Linear discriminant analysis effect size (LEfSe) was used to screen for intergroup differential marker species, by linear discriminant analysis (LDA) score >2 and P<0.05, and Kyoto encyclopedia of genes and genomes (KEGG) and MetaCyc functional prediction analyses were also performed.

Significant differences were observed among the 3 groups in terms of age, gender, body mass index (BMI), alanine aminotransferase (ALT), fasting plasma glucose (FPG) and free fatty acids (FFA) (P<0.05). There were totally 48 236 ASVs obtained, with CHB complicated with NAFLD group having the highest number of unique ASVs. Compared to the healthy control group, the CHB complicated with NAFLD group exhibited an increasing trend in species richness, community heterogeneity, and distribution uniformity of the tongue coating microbiota. At the phylum level, the proportions of Firmicutes (32%) and Proteobacteria (26%) were increased, while the proportion of Bacteroidota (27%) was decreased in the CHB complicated with NAFLD group than the healthy control group. In comparison to the CHB-only group, the proportion of Bacteroidota was increased. At the genus level, the proportions of Veillonella (18%) and Neisseria (18%) were increased, while that of Prevotella_7 (15%) was decreased in the CHB complicated with NAFLD group than the healthy control group, and that of Neisseria and Prevotella_7 was increased in the CHB complicated with NAFLD group when compared with the CHB-only group. LEfSe analysis revealed that Actinobacteriota, Neisseria_elongata, and Streptococcus_oralis were potential marker species in the CHB complicated with NAFLD group. Both KEGG and MetaCyc functional prediction analyses indicated that the differentially expressed functions of the tongue coating microbiota among the 3 groups were primarily enriched in pathways such as cofactor and vitamin metabolism, carbohydrate metabolism, and lipid metabolism.

CHB complicated with NAFLD patients exhibit characteristic changes in their tongue coating microbiota. The flora of tongue coating may affect the glucose and lipid metabolism of CHB patients through the "Oral-gut-liver axis", and promote the process and progress of NAFLD.

To systematically assess the disease burden of ischemic stroke (IS) attributable to high fasting plasma glucose (HFPG) among Chinese adults aged 25 years and older from 1990 to 2021 based on 2021 Global Burden of Disease (GBD) data, and to analyze population disparities and project trends for 2022 to 2035, providing evidence for prevention and control strategies.

Data were extracted from the GBD 2021 database, including age-standardized mortality rate (ASMR), disability-adjusted life years per million (ASDR), and population subgroup data for HFPG-related IS in China from 1990 to 2021. Joinpoint regression model was used to analyze temporal trends, and Bayesian age-period-cohort (BAPC) model was employed to predict future burden changes.

Overall, both the ASMR and ASDR for HFPG-related IS showed a declining trend from 1990 to 2021 among individuals aged 25 and older in China. However, in 2021, the ASMR (19.76 per 100 000) and ASDR (354.37 per 100 000 person-years) remained significantly higher than the global levels (ASMR=14.69 per 100 000; ASDR=266.36 per 100 000 person-years). The disease burden was higher in males than in females, with an increasing trend among those aged 25 to 44 years and a non-significant decline among those aged 80 years and older. Projections indicated that China’s ASMR and ASDR will be in a continuous decline from 2022 to 2035, with a projected rate of decrease exceeding the global average, though absolute levels will still remain higher than the global average.

Although the overall burden of HFPG-related is among Chinese individuals aged ≥25 years has decreased, the prevention and control situation remains severe. Countermeasures　It is recommended to enhance blood glucose screening for high-risk populations and standardized primary care management. Precise intervention strategies should be formulated for males, the elderly, and other young high-risk groups. Additionally, it is essential to improve the continuum of stroke treatment, rehabilitation services, and medical insurance coverage, while establishing a monitoring, evaluation, and early warning system to curb the burden of hyperglycemia-related ischemic stroke.

To investigate the association between Framingham steatosis index (FSI) and the risk of gallstone disease in order to evaluate its potential as a risk marker and simple screening tool.

A cross-sectional study was conducted on 5 705 adults subjected from the 2017-2020 cycle of the US National Health and Nutrition Examination Survey (NHANES). A complex sampling-weighted multivariable logistic regression model was employed to assess the association between FSI (treated as both a continuous variable and in quartiles) and gallstone risk. Restricted cubic spline (RCS) and threshold effect analyses were utilized to explore nonlinear relationships. Interaction and subgroup analyses were performed to examine potential effect modifiers, and multiple sensitivity analyses were applied to verify the robustness of the findings.

The fully adjusted logistic regression model showed a significant positive association between FSI and gallstone risk (OR=1.31, 95%CI: 1.20 to 1.44, P<0.001). Compared with the low FSI quartile, the high quartile had a significantly increased risk (OR=3.67, 95%CI: 2.21 to 6.09, P<0.001). Restricted cubic spline analysis suggested a nonlinear relationship between FSI and gallstone disease (P for nonlinear=0.031), with an inflection point at -0.27. Sensitivity analyses confirmed the robustness of the positive association.

Elevated FSI levels are significantly positively correlated with the risk of gallstone disease, suggesting its potential as a convenient metabolic composite indicator in community and clinical settings for the rapid identification of high-risk individuals.　Countermeasures　It is recommended that public health authorities should consider taking FSI into screening program for adult gallstone risk. Following prospective validation in Chinese populations, its threshold value could be utilized for rapid community-based risk stratification (high-vs low-risk), thereby optimizing the allocation of screening resources. Furthermore, comprehensive guidelines targeting body weight, blood lipids, blood pressure, and blood glucose should be developed and promoted. Strengthening public education and enhancing surveillance should be conducted to achieve the ultimate goals of early warning, precise intervention, and reduction of disease burden for high-risk populations.

To analyze the epidemiological trends and disease burden in chronic hepatitis B (including cirrhosis) from 1990 to 2021 globally and in China based on Global Burden of Disease Database (GBD) 2021 study, assess development-related inequality patterns and demographic drivers, and project China’s future burden and risk-factor contributions.

A secondary analysis was conducted on GBD 2021 data for 1990—2021 to extract the age-standardized incidence rate (ASIR), age-standardized prevalence rate (ASPR), and age-standardized disability-adjusted life-year rate (ASDR) globally (204 countries/territories) and in China, each with 95% uncertainty interval (UI). Joinpoint regression was used to estimate the average annual percent change (AAPC) with 95% confidence interval (CI). Combining sociodemographic index (SDI)-related inequality metrics, frontier analysis, and Das Gupta symmetric decomposition, differences in disease burden were quantified, and the relative contributions of population growth, population aging, and changes in age-specific rates to changes in absolute burden were also analyzed. Autoregressive integrated moving average (ARIMA) models were fitted to China’s annual series for 1990—2021 to forecast trends for 2022—2035. Within the GBD comparative risk assessment framework, the attributable burden from alcohol consumption and drug use were quantified.

Global ASIR, ASPR and ASDR in chronic hepatitis B were declined overall from 1990 to 2021 (all AAPCs, P<0.05). However, under the combined effects of population growth and aging, absolute numbers of cases and disability-adjusted life years (DALYs) did not decrease in parallel with age-standardized rates. The burden was concentrated primarily in low-middle SDI countries/territories in Asia and sub-Saharan Africa. In China, all 3 age-standardized indicators were decreased by more than 50% when compared with the values in 1990, and the burden was declined markedly in children; Nonetheless, middle-aged and old men remained the main high-burden group, with total cases and DALYs staying at relatively high levels. Decomposition analyses indicated that declines in age-specific rates (epidemiologic change) were the primary driver of reductions in China’s absolute burden, whereas changes in global DALYs were more readily offset or driven by population growth and aging. ARIMA projections suggested that, under continuation of historical control intensity, China’s ASIR and ASDR will decline slowly through 2035, while ASPR will remain comparatively high. Comparative risk assessment showed that alcohol consumption and drug use were important modifiable contributors to DALYs related to chronic hepatitis B (including cirrhosis) globally and in China; alcohol-attributable ASDR in China was higher than the global average.

From 1990 to 2021, the age-standardized burden of chronic hepatitis B (including cirrhosis) declined globally and in China, yet substantial absolute-burden pressure persists worldwide due to population growth and aging and is unevenly distributed. Despite major progress in China in vaccination, prevention of mother-to-child transmission, and antiviral therapy, strengthened management of prevalent infections and prevention of complications remain critical to further reduce residual burden. Countermeasures　In line with the 2030 goal of eliminating viral hepatitis as a public health threat of World Health Organization, we recommend sustaining high coverage of timely birth-dose and full-course hepatitis B vaccination and strengthening prevention of mother-to-child transmission (PMTCT); prioritizing middle-aged and old men and high-risk areas to improve coverage of screening, follow-up, and treatment; integrating alcohol reduction and services for substance use disorders into hepatitis B care; and directing testing and treatment resources toward low-middle SDI settings and other resource-limited areas to narrow burden disparities.

To investigates the association between the variability of remnant cholesterol (RC) during long-term follow-up and cardiac conduction block (CCB) based on the Kailuan cohort.

A retrospective study was conducted on the research cohort in the Kailuan study population who consecutively participated in the physical examinations of 2006, 2008, and 2010. Based on the inclusion and exclusion criteria, 39 299 subjects were ultimately enrolled in the study. The variability of RC during the follow-up period was calculated, with the variability assessed through standard deviation (SD), average real variability (ARV), coefficient of variation (CV), and variability independent of the mean (VIM). The participants were divided into 4 groups based on the interquartile range of RC variability, and followed up every 2 years. The endpoints were defined as the occurrence of CCB, death, or the termination of follow-up (December 31, 2019). Kaplan-Meier test and Log-Rank test were carried out respectively calculate and compare the cumulative incidence rate of endpoint events across groups. Cox proportional hazards model and restricted cubic spline (RCS) analysis were performed to analyze the association between the variability of RC and CCB and its subtypes.

After an average follow-up of 7.43 years in 39 299 participants, 916 patients developed CCB. After adjusting for confounding factors, using CCB as the outcome event, and were dividing the participants into groups based on the RC-SD quartiles, the hazard ratios (HR) for CCB occurrence were 1.344 in the Q1 group (95% CI: 1.110~1.626, P=0.002), 1.068 (95% CI: 0.878 to 1.299, P=0.509) in the Q3 group and 1.311 (95% CI: 1.089~1.579, P=0.004) in the Q4 group, when compared to the group of lowest incidence density (the Q2 group). Similar results were obtained for the quartiles of RC-ARV, RC-CV, and RC-VIN grouping. RCS analysis revealed a U-shaped association between RC-SD and CCB (P_overall<0.05, P_nonlinearity<0.05). Sensitivity and stratified analyses further validated this finding.

There exists a U-shaped association between the variability of RC and the risk of CCB. This finding provides a new direction for preventing the occurrence of CCB. Countermeasures　China’s blood lipid management strategy should undergo a paradigm shift from pursuing “low-level stability” toward integrating “long-term stability of biomarker levels” This entails systematically incorporating the new concept of “maintaining physiological fluctuations of blood lipids within an optimal range” into the national prevention and treatment plans for cardiovascular and cerebrovascular diseases, primary public health service programs, and the public health education system. Such a transition will propel chronic disease management into a more precise and scientific new stage.

To investigate the risk factors and causal relationships of coronary heart disease (CHD) by utilizing clinical data and genetic variation data, and construct an interpretable risk prediction model for the disease.

A retrospective cohort analysis was conducted on the clinical data and genetic variation data collected from the Kaggle platform and the IEU OpenGWAS Project database, respectively. SMOTEENN mixed sampling was used to balance the data. Logistic regression models and two-sample Mendelian randomization (MR) were combined to screen risk factors and analyze their causal effects. Then risk prediction models were further constructed using random forest (RF), eXtreme gradient boosting (XGBoost), and artificial neural networks (ANN), and feature importance was assessed with the SHAP algorithm.

Logistic regression analysis showed that gender, age, education level, smoking, hypertension, total cholesterol (TC), systolic blood pressure (SBP), and blood glucose were significantly associated with CHD (P<0.05), while body mass index (BMI), diastolic blood pressure (DBP), history of stroke, and diabetes had no such association. Two-sample MR analysis indicated that BMI (OR=1.014, 95%CI: 1.011 to 1.017, P=1.720×10^-22), age (OR=1.015, 95%CI: 1.012 to 1.019, P=3.455×10^-18), number of cigarettes smoked per smoker per day (OR=1.007, 95%CI: 1.002 to 1.011, P=0.005), hypertension (OR=1.057, 95%CI: 1.046 to 1.067, P=5.870×10^-28) were positively associated with CHD; Education level (OR=0.970, 95%CI: 0.960 to 0.979, P=1.231×10^-9) and high-density lipoprotein (OR=0.990, 95%CI: 0.986 to 0.993, P=0.002) were negatively associated with CHD. In the machine learning models, the SMOTEENN-RF performed best, with an accuracy rate of 0.887 and an AUC value of 0.962. SHAP analysis indicated that age, gender, hypertension, SBP, and cigarettes smoked per day contributed most to the model.

Key risk factors such as age, hypertension, and blood lipid levels are significantly correlated with CHD. A high-performing SMOTEENN-RF prediction model is successfully constructed, providing reliable methodological support and practical tool for CHD risk early warning and precision prevention and treatment.

To explore the transition patterns of latent classes of depressive symptoms among postgraduate students and their influencing factors.

A prospective cohort study was conducted. Two waves of survey were carried out on 5 577 postgraduate students from an university in Hubei Province and another one in Anhui Province in October 2021 (T1) and October 2022 (T2). Self-Rating Depression Scale and College Students Mental Health Screening Scale were applied respectively to measure the severities of depressive symptoms and developmental distress. The optimal classification model was determined through the fitting indicators of the latent class model, mainly including indicators such as Entropy value, Lo-Mendell-Rubin (LMR), and Bootstrapped Likelihood Ratio Test (BLRT). Latent transition analysis was used to obtain the transition probabilities of latent class changes at different time points, and multivariate logistic regression analysis was employed to identify the factors influencing the latent class transition of depressive symptoms in the postgraduate students.

① The depressive symptoms could be classified into 3 categories among the postgraduate students (T1: Entropy=0.85, LMRP<0.05, BLRTP<0.001; T2: Entropy=0.88, LMRP<0.001, BLRT P<0.001), that is, a non-depressive symptom group (66.0% vs 64.3%), a mild depressive symptom group (25.6% vs 26.9%), and a moderate depressive symptom group (8.4% vs 8.8%). ② The probabilities of each latent class remaining in the original group at time point T2 were 84.1%, 56.9%, and 26.6%, respectively. The probability of transition from moderate to mild depressive symptom group was 41.3%, and the probability from mild to non-depressive symptom group was 29.3%.③ Adaptation challenges(OR=1.08, 95%CI: 1.02 to 1.16, P<0.05) and employment stress (OR=1.05, 95CI: 1.01 to 1.10, P<0.05) were promoting factors for the transition from non-depressive to mild depressive symptom group. Romantic troubles (OR=0.86, 95%CI: 0.76 to 0.99, P<0.05) was an obstructive factor for the transition from moderate to non-depressive symptom group.

There are 3 latent categories of depressive symptoms among postgraduate students. an obstructive factor for the transition from moderate to non-depressive symptom group.Adaptation challenges, employment stress and romantic troubles are important factors for the development of depressive symptoms.