Xiaobing Zhai

Background: Obesity is a well-established risk factor for major depressive disorder (MDD), yet the risk is not uniform, highlighting the need for precise risk stratification. This study aimed to develop a metabolomics-based prediction model to identify high-risk metabolic phenotypes among obese participants and to elucidate the causal metabolic pathways involved. Methods: Forty-one-thousand-four-hundred-fifty-nine obese participants were followed for a median of 14.4 years. We integrated multiple machine learning (ML) algorithms to develop predictive models for 3-, 5-, and 9-year MDD risk, with a temporal validation within the same biobank. Furthermore, we investigated the potential causal relationships within the obesity-metabolite-MDD using mediation Mendelian randomization (MR). Results: During follow-up, 3,642 incident MDD cases were documented. The optimized LightGBM model demonstrated superior predictive performance, achieving AUCs of 0.844 (95% CI: 0.773–0.914), 0.824 (95% CI: 0.771–0.875), and 0.834 (95% CI: 0.796–0.871) for 3-, 5-, and 9-year intervals, respectively, significantly outperforming existing clinical models. Temporal validation confirmed the model’s robustness (AUCs: 0.738–0.776). MR analysis confirmed that key metabolites causally mediate the pathway from obesity to MDD (mediation proportions: −11.5% and −35.3%, all P _ME < 0.05). Conclusions: These findings challenge the notion of a uniform obesity–MDD association, demonstrating that metabolomic signatures can effectively stratify MDD risk. We present a validated ML framework for the early identification of high-risk individuals with obesity, offering a precision medicine approach to guide targeted metabolic and psychiatric interventions.

Background: Metabolic dysfunction-associated fatty liver disease (MAFLD) affects 38.9% of the global adult population and is associated with increased mortality when it co-occurs with depression. Women exhibit a 1.5– to 3–fold higher prevalence of depression, with postmenopausal hormonal imbalances amplifying susceptibility. This underscores the urgent need for sex-specific predictive models. The aim of this study was to develop a lightweight, high-accuracy model to identify key predictors of depression risk in female MAFLD patients using a five-year longitudinal UK Biobank cohort. Methods: We analyzed routine blood biomarkers (hematology and metabolites), lifestyle factors, and reproductive history from female participants with MAFLD identified within the UK Biobank cohort. Logistic regression was adjusted for socioeconomic, lifestyle, multimorbidity, and medication use, and was employed to evaluate sex-specific factors. Feature selection employed a two-stage approach to ensure balanced covariate distributions between cases and controls: Random Forest-based stratified bootstrap resampling (1000 iterations) instead of traditional random sampling, followed by recursive feature elimination. Five models—Light Gradient Boosting Machine (LightGBM), Extreme Gradient Boosting (XGBoost), Random Forest, Feature Tokenizer Transformer (FT-Transformer), and Gated Adaptive Network for Deep Automated Learning of Features (GANDALF)—were assessed via 10–fold cross-validation. SHapley Additive exPlanations (SHAP) and restricted cubic splines (RCS) elucidated feature importance and nonlinear effects. Results: A total of 39,430 female MAFLD patients were included in the final analysis, among whom 611 (1.55%) developed incident depression over the five-year follow-up. Adjusted logistic regression identified younger age at first live birth (30 kg/m2), and early reproductive milestones. Additionally, our feature set showed enhanced predictive validity compared to established biomarker panels from prior studies (ROC-AUC 0.96 vs. 0.886–0.940). Conclusions: This study introduces a highly accurate, lightweight predictive model tailored for female MAFLD patients, leveraging 17 key features to improve the prediction of depression risk. By enabling personalized risk assessment and targeted interventions, our model offers a transformative approach to improve mental health outcomes and care quality in this vulnerable population.

Background: Incorporating initial serum chloride levels as a prognostic indicator in the intensive care environment has the potential to refine risk stratification and tailor treatment strategies, leading to more efficient use of clinical resources and improved patient outcomes. Objective: Quantitative analysis of the relationship between serum chloride levels at intensive care unit (ICU) admission and in-hospital mortality, and the establishment of a personalized survival curve prediction deep learning model to enhance risk stratification and clinical decision-making. Methods: A large-scale, cross-country, multicohort study of 189,462 ICU patients from four cohorts was conducted: 70,370 from Medical Information Mart for Intensive Care IV (MIMIC-IV), 112,457 from eICU Collaborative Research Database (eICU-CRD; 2 US cohorts), 4653 from Yantai Yuhuangding Hospital, and 1982 patients from Zigong Fourth People’s Hospital (2 Chinese cohorts). We collected demographics, underlying diseases, ICU complications, electrolyte levels, biochemical parameters, and vital signs at ICU admission, along with length of stay and in-hospital survival outcomes. Causal graph analysis pinpointed clinical variables linked to mortality. Nonlinear associations between chloride levels and mortality were evaluated using restricted cubic splines and Cox proportional hazards models, validated with the Cox frailty model, Kaplan-Meier curves, and sensitivity analyses. A deep learning model was created for individualized survival predictions. Results: Causal inference revealed a significant association between admission serum chloride levels and 28-day mortality. The median serum chloride level at ICU admission was 104 (IQR 100-108) mEq/L. In analyzing all 42 variables, restricted cubic splines identified thresholds at 103 mEq/L and 115 mEq/L, categorizing patients into three groups: ≤103 mEq/L, 103-115 mEq/L, and >115 mEq/L. Cox proportional hazards models revealed higher death risks for patients outside this range, with hazard ratios (HRs) of 1.36 (95% CI 1.29-1.43) for ≤103 mEq/L and 1.27 (95% CI 1.14-1.41) for >115 mEq/L. Four cross-cohort validations confirmed these critical ranges. For the eICU-CRD dataset, the HRs for the key intervals are 1.30 (95% CI 1.24-1.36) and 0.97 (95% CI 0.89-1.06). In the Yantai Yuhuangding Hospital affiliated with Qingdao University (YHD-HOSP) dataset, the HRs for the key intervals are 1.23 (95% CI 1.09-1.38) and 1.58 (95% CI 1.27-1.96). In the Sichuan Zigong Fourth People’s Hospital (SCZG-HOSP) dataset, the HR for the key interval is 2.20 (95% CI 1.43-3.39). The Causal SurvivalNet accurately predicted individual survival curves using admission chloride levels and other factors, achieving Brier scores of 0.09, 0.12, and 0.15. Results from cohort analyses in both China and the United States consistently and closely correlate the critical range of chloride with the prognosis of ICU patients. Conclusions: Using initial serum chloride levels enhances prognostic accuracy and facilitates tailored treatment plans for ICU patients in critical care settings.

Background: To investigate the relationship of age at first separation from parents (AFSP) with suicidal ideation (SI), suicide planning (SP) and suicide attempts (SA) among left-behind children (LBC) in China and the mediating effect of traumatic events (TEs). Methods: This cross-sectional study employed a multistage sampling method to conduct questionnaire surveys in six high-concentrated LBC provinces. Logistic regression and restricted cubic splines were performed to investigate the association of AFSP with SI, SP and SA, and meta-analysis (I²) was used to assess the heterogeneity. Causal mediation analysis (structural equation modelling) was used to evaluate the mediating effect of TEs, and bootstrap resampling was employed to assess the robustness of the causal mediation analysis. Results: A total of 12 162 (6269 LBC) participants were included in this study; the prevalence rates of SI, SP and SA among LBC were 30.7%, 18.6% and 11.9%; and LBC had a higher risk of SI (OR: 1.32), SP (OR: 1.40) and SA (OR: 1.25) than non-LBC, all p<0.001. AFSP was negatively associated with SI, SP and SA, particularly peaking before 2 years old. LBC with AFSP <2 years old had a higher risk of SI (OR 2.65; 95% CI 2.17 to 3.24; p<0.001), SP (OR 2.63; 95% CI 3.36 to 4.29; p<0.001) and SA (OR 2.04; 95% CI 2.73 to 3.63; p<0.001). TEs mediated the increased risk of SI (average causal mediation effect (ACME) 0.03–0.07; proportion (Prop) 40–54%; all p<0.001), SP (ACME 0.03–0.06; Prop 39–61%; all p<0.001) and SA (ACME 0.02–0.04; Prop 53–80%; all p<0.01). Conclusions: We found that AFSP was associated with SI, SP and SA, mediated by TEs. Our findings supported the necessity of reducing SI, SP and SA risks among LBC and highlighted the need to consider AFSPs and TEs in intervention strategies.

BACKGROUND Incorporating initial serum chloride levels as a prognostic indicator in the intensive care environment has the potential to refine risk stratification and tailor treatment strategies, leading to more efficient use of clinical resources and improved patient outcomes. OBJECTIVE Quantitative analysis of the relationship between serum chloride levels at intensive care unit (ICU) admission and in-hospital mortality, and the establishment of a personalized survival curve prediction deep learning model to enhance risk stratification and clinical decision-making. METHODS A large-scale, cross-country, multicohort study of 189,462 ICU patients from four cohorts was conducted: 70,370 from Medical Information Mart for Intensive Care IV (MIMIC-IV), 112,457 from eICU Collaborative Research Database (eICU-CRD; 2 US cohorts), 4653 from Yantai Yuhuangding Hospital, and 1982 patients from Zigong Fourth People’s Hospital (2 Chinese cohorts). We collected demographics, underlying diseases, ICU complications, electrolyte levels, biochemical parameters, and vital signs at ICU admission, along with length of stay and in-hospital survival outcomes. Causal graph analysis pinpointed clinical variables linked to mortality. Nonlinear associations between chloride levels and mortality were evaluated using restricted cubic splines and Cox proportional hazards models, validated with the Cox frailty model, Kaplan-Meier curves, and sensitivity analyses. A deep learning model was created for individualized survival predictions. RESULTS Causal inference revealed a significant association between admission serum chloride levels and 28-day mortality. The median serum chloride level at ICU admission was 104 (IQR 100-108) mEq/L. In analyzing all 42 variables, restricted cubic splines identified thresholds at 103 mEq/L and 115 mEq/L, categorizing patients into three groups: ≤103 mEq/L, 103-115 mEq/L, and >115 mEq/L. Cox proportional hazards models revealed higher death risks for patients outside this range, with hazard ratios (HRs) of 1.36 (95% CI 1.29-1.43) for ≤103 mEq/L and 1.27 (95% CI 1.14-1.41) for >115 mEq/L. Four cross-cohort validations confirmed these critical ranges. For the eICU-CRD dataset, the HRs for the key intervals are 1.30 (95% CI 1.24-1.36) and 0.97 (95% CI 0.89-1.06). In the Yantai Yuhuangding Hospital affiliated with Qingdao University (YHD-HOSP) dataset, the HRs for the key intervals are 1.23 (95% CI 1.09-1.38) and 1.58 (95% CI 1.27-1.96). In the Sichuan Zigong Fourth People’s Hospital (SCZG-HOSP) dataset, the HR for the key interval is 2.20 (95% CI 1.43-3.39). The Causal SurvivalNet accurately predicted individual survival curves using admission chloride levels and other factors, achieving Brier scores of 0.09, 0.12, and 0.15. Results from cohort analyses in both China and the United States consistently and closely correlate the critical range of chloride with the prognosis of ICU patients. CONCLUSIONS Using initial serum chloride levels enhances prognostic accuracy and facilitates tailored treatment plans for ICU patients in critical care settings.

Depression represents a significant global public health challenge, and marital status has been recognized as a potential risk factor. However, previous investigations of this association have primarily focused on Western samples with substantial heterogeneity. Our study aimed to examine the association between marital status and depressive symptoms across countries with diverse cultural backgrounds using a large-scale, two-stage, cross-country analysis. We used nationally representative, de-identified individual-level data from seven countries, including the USA, the UK, Mexico, Ireland, Korea, China and Indonesia (106,556 cross-sectional and 20,865 longitudinal participants), representing approximately 541 million adults. The follow-up duration ranged from 4 to 18 years. Our analysis revealed that unmarried individuals had a higher risk of depressive symptoms than their married counterparts across all countries (pooled odds ratio, 1.86; 95% confidence interval (CI), 1.61–2.14). However, the magnitude of this risk was influenced by country, sex and education level, with greater risk in Western versus Eastern countries (β = 0.36; 95% CI, 0.16–0.56; P < 0.001), among males versus females (β = 0.25; 95% CI, 0.003–0.47; P = 0.047) and among those with higher versus lower educational attainment (β₂ = 0.34; 95% CI, 0.11–0.56; P = 0.003). Furthermore, alcohol drinking causally mediated increased later depressive symptom risk among widowed, divorced/separated and single Chinese, Korean and Mexican participants (all P < 0.001). Similarly, smoking was as identified as a causal mediator among single individuals in China and Mexico, and the results remained unchanged in the bootstrap resampling validation and the sensitivity analyses. Our cross-country analysis suggests that unmarried individuals may be at greater risk of depression, and any efforts to mitigate this risk should consider the roles of cultural context, sex, educational attainment and substance use.

Objectives: Evidence of the relationshipbetween android fat mass and gynoid fat mass with the mortality prediction is still limited. Current study analyzed the NHANES database to investigate the relationship between android fat mass, gynoid fat mass and CVD, with all-cause mortality.Method: The study subjects were NHANES participants over 20 years old, two indicators of regional body composition, android fat and gynoid fat were measured by Dual Energy x-ray Absorptiometry (DEXA). The other various covariates data obtained from the NHANES questionnaire and laboratory measurements, including age, gender, education, race/ethnicity, uric acid, total serum cholesterol, albumin, Vitamin C, folate, alcohol drinking, smoking status, history of diabetes, and hypertension. Mortality status was ascertained from a linked mortality file prepared by the National Center for Health Statistics. The study population was divided quartiles based on the distribution of android fat mass and gynoid fat mass. The relationship between these two indicators with cardiovascular and all-cause mortality was investigated by using Cox regression. The covariates age, gender, smoking status, drinking status, history of diabetes, and history of hypertension were stratified.Results: In the fully adjusted model, Q3 had the lowest HR in android fat mass and gynoid fat mass. When examining the relationship between android fat mass and CVD mortality, current smokers and drinkers had the lowest CVD risk in Q2 [smoking: 0.21 (0.08, 0.52), drinking: 0.14 (0.04, 0.50)]. In diabetic patients, compared with Q1, other groups with increased android fat mass can significantly reduce the risk of CVD [Q4: 0.17 (0.04, 0.75), Q3: 0.18 (0.03, 1.09), Q2: 0.27 (0.09, 0.83)]. In ≥60 years old and female, the greater the gynoid fat mass, the smaller the HR of all-cause mortality [Q4 for ≥60 years old: 0.57 (0.33, 0.96), Q4 for female: 0.37 (0.23, 0.58)]. People <60 years old had a lower risk of all-cause mortality with gynoid fat mass in Q3 than those ≥60 years old [<60 years: 0.50 (0.27, 0.91), ≥60 years: 0.65 (0.45, 0.95)]. Among subjects without hypertension, the group with the largest android fat mass had the lowest risk of CVD mortality, and the group with the largest gynoid fat mass had the lowest risk of all-cause mortality [Android fat mass: 0.36 (0.16, 0.81), gynoid fat mass: 0.57 (0.39, 0.85)].Conclusion: Moderate android fat mass and gynoid fat mass (Q3) had the most protective effect. Smokers and drinkers need to control their body fat. Being too thin is harmful to people with diabetes. Increased gynoid fat mass is a protective factor for all-cause mortality in older adults and females. Young people's gynoid fat mass is more protective in the moderate range than older people's. If no high blood pressure exists, people with more android and gynoid fat mass have a lower risk of CVD or all-cause mortality.

Among patients with diabetes mellitus, limited studies have investigated the relationship between anthropometric parameters and cardiovascular disease (CVD), with all-cause mortality. We examined the associations of arm circumference (AC), arm-to-waist ratio (AC/WC), and CVD, with all-cause mortality among patients with diabetes. This is a cohort study of 5497 diabetic individuals aged 20 or over who were recruited in the National Health and Nutrition Examination Survey (NHANES) from 1999 to 2014. Cox proportional hazards regression models were used to analyze the associations between AC, AC/WC, and CVD, with all-cause mortality. We also conducted stratified analyses and explored the possible non-linear relation by restricted cubic splines. During a median follow-up of 7.2 years, there were 271 and 1093 cases of CVD and all-cause death. The multivariable adjusted hazard ratios (HRs) with 95% confidence intervals (Cis) of CVD mortality in Q2, Q3, and Q4 groups compared with Q1 group were 0.37 (0.22, 0.62), 0.24 (0.12, 0.48), 0.18 (0.07, 0.46) for AC, and 0.18 (0.07, 0.46), 0.34 (0.20, 0.60), 0.28 (0.15, 0.53) for AC/WC. Similar results were observed in the analysis for all-cause mortality risk. AC and AC/WC were both inversely associated with CVD and all-cause mortality among individuals with diabetes. It is important to pay attention to these anthropometric parameters of diabetic patients.

In the general population, there is little evidence of a link between blood urea nitrogen (BUN) and long-term mortality. The goal of this study was to explore whether higher BUN concentration is a predictor of cardiovascular disease (CVD) and all-cause mortality. From 1999 to 2006, the National Health and Nutrition Examination Survey (NHANES) included 17,719 adult individuals. Death outcomes were ascertained by linkage to the database records through 31 December 2015. The Cox proportional hazard regression model was used to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) for CVD and all-cause mortality in individuals. We also performed stratified analyses based on age, gender, drinking, smoking, history of hypertension and diabetes. During a mean follow-up 11.65 years, a total of 3628 deaths were documented, of which 859 were due to CVD. Participants with higher BUN had a higher risk of CVD and all-cause death compared to those with lower BUN. After multifactor adjustment for demographics, major lifestyle factors, and hypertension and diabetes history, higher BUN levels compared with lower levels were significantly associated with higher risk of CVD (HR: 1.48 [1.08, 2.02], P-trend < 0.001) and all-cause mortality (HR: 1.48 [1.28, 1.72], P-trend < 0.001). In subgroup analyses, we found that the trend in the association of BUN with the risk of death remained strong in female subjects. Greater BUN levels were linked to higher CVD and all-cause mortality in the NHANES of American adults. The importance of BUN in predicting death is supported by our research.