This study aimed to evaluate the effect of key social environmental determinants on trends in depression burden amongst Chinese adolescents aged 10-24 years and to examine the age- and gender-specific variations and lagged effects of these factors. Nationwide data spanning 2003-2021 were collected from the Global Burden of Disease database for depression burden indicators, World Bank socioeconomic datasets and the Chinese General Social Survey for education and social metrics. The primary variables analysed were gross domestic product (GDP) per capita, higher education enrolment rate (HEER), per capita current health expenditure and urban population percentage. Descriptive statistics and Pearson correlations were used to explore variable distributions and associations. Mixed-effects regression models quantified relationships between social factors and depression burden, and autoregressive distributed lag models assessed short-term lagged effects across age groups(10-14, 15-19 and 20-24 years). Interaction terms (HEER × total public expenditure on education (TPEE); unemployment × dependency ratio) were included on the basis of theoretical and empirical support from prior studies. Between 2003 and 2021, the overall burden of depression amongst young Chinese people decreased, but the 10- to 14-year age group showed a notable upward trend in disability‑adjusted life‑years since 2017. High GDP per capita, urbanisation and per capita health expenditure were significantly associated with reduced depression burden (p < 0.05). Conversely, increasing HEER- indicative of rising educational competition-was associated with a high disease burden, particularly amongst females aged 15-19 and 20-24 years. Short-term lagged effects revealed that the psychological burden of education competition manifested within 1 year, particularly amongst younger adolescents. Economic improvements and health investments exerted protective lagged effects. Age-stratified analyses underscored distinct vulnerability patterns: 10- to 19-year-olds were highly sensitive to family and educational support, and 20- to 24-year-olds were greatly affected by social structural pressures. Pearson correlation analysis identified significant negative associations between several social factors and depression burden. The mechanisms were explained by China-specific social contexts and data characteristics. This nationwide longitudinal study reveals that multidimensional social determinants exert age- and gender-specific influences on depression burden amongst young Chinese people aged 10-24 years. The findings emphasise the urgent need for stratified policies, including enhancing family and educational support for individuals under 20 years old and reducing structural social pressures on young adults. Public mental health interventions should target these modifiable social determinants to reduce the depression burden and improve well-being.
This study aimed to investigate the association between borderline personality traits and addictive features of non-suicidal self-injury (NSSI) in adolescent patients with depressive disorder, identify independent risk factors for NSSI addictive features and provide evidence for clinical intervention. A cross-sectional study design was utilised. A total of 320 adolescent patients with depressive disorder (aged 12-18 years, mean ± SD: 15.82 ± 1.74 years; 238 females, 74.38%) admitted to Jingzhou Mental Health Center between January 2024 and October 2025 were enrolled. Assessments were conducted using the 17-item Hamilton Depression Rating Scale, the Borderline Personality Features Scale for Children (BPFS-C), the Adolescent Self-Rating Life Events Checklist (ASLEC), the 20-item Toronto Alexithymia Scale, the Family APGAR Index and the Addiction Subscale of the Ottawa Self-Injury Inventory (OSI-AS). Pearson correlation analysis was conducted to explore the association between borderline personality traits and NSSI addictive features. Binary logistic regression analysis was performed to identify independent influencing factors for NSSI addictive features. Based on the presence or absence of NSSI addictive features, the 201 patients in the NSSI group were further divided into an addictive NSSI subgroup (n = 111) and a non-addictive NSSI subgroup (n = 90). The OSI-AS score was significantly higher in the addictive NSSI subgroup than in the non-addictive subgroup (p < 0.001). The total BPFS-C score and its subscale scores showed significant positive correlations with NSSI addictive features (p < 0.001). Binary logistic regression analysis revealed that, after adjusting for confounders such as age, gender, severity of depression and family function, the total BPFS-C score (odds ratio [OR] = 1.116, 95% confidence interval [CI]: 1.077-1.157, p < 0.001) and the total ASLEC score (OR = 1.051, 95% CI: 1.021-1.082, p = 0.001) were independent risk factors for NSSI addictive features. The overall prediction accuracy of this model was 84.7%. Borderline personality traits are an independent risk factor for NSSI addictive featuresin adolescent patients with depressive disorder and are closely associated with the severity of addictive NSSI. In clinical practice, screening for borderline personality traits should be implemented for adolescents with depression and NSSI. Early psychological interventions targeting core features such as affective instability and impulsivity should be conducted to reduce the risk of NSSI addiction.
Adolescent depression is an increasing public health concern, with excessive screen time elevating depression risk and activity interests providing protection. However, most studies examine these behaviors separately and rely on limited analytical methods. This study used machine learning (ML) to develop a predictive model and evaluate the combined influence of screen time and activity interests on adolescent mental health. A multi-center survey was conducted among adolescents aged 10-14 years in Chongchuan District, Nantong. Depression-related domains were assessed using the Child and Adolescent Mental Health Screening Questionnaire, integrating seven validated scales. A twostage feature-selection strategy identified 11 key predictors. Three ML models (logistic regression [LR], extreme gradient boosting [XGBoost], and categorical boosting [CatBoost]) were trained with an 80:20 stratified split. Class imbalance was addressed using synthetic minority oversampling technique and class-weighting. Model performance and interpretability were evaluated using receiver operating characteristic (ROC) and calibration curves, partial dependence plots, and shapley additive explanations (SHAP) analyses. A total of 2202 valid questionnaires were analyzed. The distribution of depression severity was as follows: safe 59%, mild 17%, moderate 11%, and severe 13%. The integrated questionnaire demonstrated strong reliability (Cronbach's α = 0.910) and good construct validity (Kaiser-Meyer-Olkin [KMO] = 0.91; root mean square error of approximation [RMSEA] = 0.049; and comparative fit index [CFI] = 0.859). ROC-Youden analysis confirmed expert-defined cutoffs (29, 32, and 35). Feature selection identified 11 key predictors, with activity interest and psychological functioning consistently ranking highest in importance. Across the three ML models, LR exhibited the best generalizability, XGBoost showed overfitting, and CatBoost achieved balanced performance. SHAP and partial dependence analyses revealed nonlinear screen-time effects and dose-dependent protective effects of activity interest, including the moderation of high screen exposure in severe-risk groups. This study suggests that ML models can be used to screen adolescents at risk of depression by capturing the combined influence of screen time and activity interests. The model is intended for screening rather than diagnosis and may support school-based early identification, and further validation in clinical contexts is needed.
Self-harm, which includes both nonsuicidal self-injury and suicidal behaviors, poses a major global public health challenge. This study provides a comprehensive analysis of trends in self-harm worldwide, the socioeconomic disparities associated with it, and future projections, using data from the Global Burden of Disease, Injuries, and Risk Factors Study (GBD) 2021. Self-harm data were extracted from GBD 2021, including incidence, prevalence, mortality, years of life lost (YLLs), years lived with disability (YLDs), and disability-adjusted life years(DALYs) for 204 countries and territories from 1990 to 2021. Age-standardized rates and estimated annual percentage change (EAPC) were calculated. Inequality was assessed using the Slope Index of Inequality (SII) and Concentration Index (CI). Autoregressive Integrated Moving Average (ARIMA) models were employed to generate projections of self-harm burden from 2022 to 2040. The global burden of self-harm is projected to change substantially by 2040, with deaths estimated to increase to 829,853 (95% Uncertainty Interval (UI), 262,233-1,397,474) and prevalence projected to rise to 35,863,341 (95% UI, 8,079,108-63,647,574) cases (representing a 131.9% increase from the 2021 baseline of 15,467,153 cases). From 1990 to 2021, age-standardized rates of self-harm demonstrated decreasing trends globally and across sociodemographic index (SDI) levels, with the largest declines observed in high-middle SDI countries. Gender disparities were evident, with more pronounced decreases in females. Inequalities in DALYs due to self-harm decreased over time but remained higher among females in lower-SDI populations. Despite decreasing age-standardized rates, the global burden of self-harm is projected to increase substantially by 2040, with driven by increasing incidence and prevalence in incidence and prevalence. Inequities persist, particularly among females in lower-SDI populations. Implementation of targeted prevention and intervention strategies, strengthening of mental health systems, and addressing social determinants of health are imperative to reduce the growing burden of self-harm worldwide.
Fetal loss constitutes a major obstetric adverse outcome, and is frequently followed by marked psychological distress; the prevalence of depressive symptoms after fetal loss is substantially higher, and this elevation is intertwined with psychosocial determinants whose clinical profiles and intervention targets await systematic synthesis. This study examines the clinical characteristics and psychosocial determinants of maternal depression following fetal loss, aiming to inform targeted psychological support strategies. Thisretrospective study included 200 mothers following fetal loss (fetal loss group) and 200 mothers after normal delivery (term delivery group), selected via 1:1 nearest-neighbour propensity score matching (PSM) between June 2022 and October 2025. At 42 days postevent, participants in both groups completed the Edinburgh Postnatal Depression Scale (EPDS), Hospital Anxiety and Depression Scale (HADS), Multidimensional Scale of Perceived Social Support (MSPSS), and the Olson Marital Quality Questionnaire (ENRICH). Depression was defined as EPDS ≥13. Depression prevalence and scale scores were compared between groups, and multivariable logistic regression identified risk factors associated with depression. Changes in EPDS score reduction (Δ = score at 3 months postpartum - baseline score) was compared between those who received and those who did not receive clinical management. Following PSM, baseline characteristics were well-balanced between the two groups(p > 0.05). The prevalence of depression was significantly higher among women with the fetal loss than among those with term delivery group (35.0% vs. 8.50%, p < 0.05). Multivariate analysis identified fetal loss as an independent predictor of depression (odds ratio (OR) = 2.84, 95% confidence interval (CI): 1.96-4.12). EPDS scores were significantly higher in the fetal loss group than in the term delivery group (13.1 ± 4.0 vs. 8.5 ± 2.0, p < 0.001). The predominant symptoms included persistent low mood (87.1%), insomnia (75.7%), guilt or self-blame (68.6%) and fear or avoidance of future pregnancy (62.9%). Within the fetal loss group, Low social support (OR = 3.15), marital dissatisfaction (OR = 2.43), ≥2 abortions (OR = 1.98), and lack of clinical management (OR = 2.27) were independently predicted depression. Only 27.6% of affected mothers received treatment, and this was associated with significantly greater improvement in EPDS scores (△ = -5.2 ± 2.4 vs. -1.9 ± 2.0, p < 0.001). Fetal loss is associated with a substantially increased risk of maternal depression, characterized by self-blame and fear of future pregnancy. Modifiable factors including low social support and absent professional care, are associated with more persistent depressive symptoms. These Findings support the intervention integration for high-risk mothers, although further validation is required.
Primary caregivers of children with hematological malignancies endure immense physical and psychological stress, however their mental health status remains under-recognized in clinical settings. This study aimed to investigate the prevalence of anxiety and depression among these caregivers and to identify their independent associated factors, providing evidence for targeted nursing interventions. A cross-sectional survey was conducted involving 200 primary caregivers recruited from the Department of Haematology, Children's Hospital of Soochow University. Demographic and clinical data were collected alongside psychological assessments using the Hospital Anxiety and Depression Scale (HADS), the Memorial Symptom Assessment Scale (MSAS), and the Multidimensional Scale of Perceived Social Support (MSPSS). Univariate analyses and multivariable logistic regression models were employed to determine the associations between potential predictors and psychological distress outcomes. The study revealed a substantial burden of psychological morbidity. Specifically, 52.00% of caregivers exhibited anxiety and 42.50% showed symptoms of depression. Multivariable analysis further identified distinct risk profiles for each condition. Anxiety was independently associated with shorter time since diagnosis (odds ratio (OR) = 0.66, 95% confidence interval (CI): 0.47-0.92, per 6 months), greater child symptom burden (OR = 1.22, 95%: CI 1.03-1.46, per 10 points increase), and social support (OR = 0.48, 95% CI: 0.32-0.71, per 10 points). Conversely, depression was significantly associated with sociodemographic factors including educational level (OR = 0.34, 95% CI: 0.13-0.89, for college degree or above) and single marital status (OR = 3.97, 95% CI: 1.35-11.69), in addition to symptom burden (OR = 1.55, 95% CI: 1.25-1.91, per 10 points) and social support (OR = 0.31, 95% CI: 0.19-0.50, per 10 points). Furthermore, sensitivity analyses highlighted that the frequency recent hospitalizations were consistently associated with higher levels of both anxiety and depression. Caregivers of children with haematological malignancies experience exhibits a high prevalence of anxiety and depression. Anxiety appears to be more closely related to acute clinical stressors and temporal factors, whereas depression is more closely related to persistent social and demographic disadvantages. Effective management of paediatric symptoms and the strengthening of multi-dimensional social support systems are essential. Future interventions should be tailored to the specific risk profiles of caregivers to improve their overall well-being.
To investigate the prognostic influence of preoperative cognitive function, self-efficacy, and postoperative psychological counselling on treatment response following botulinum toxin injections in patients with anxiety and depressive disorders, and to identify key predictors of treatment response. A retrospective study was conducted on 176 patients who received botulinum toxin injections at Huzhou Maternity and Child Health Care Hospital between May and December 2025. Based on the treatment response of anxiety and depressive symptoms assessed eight weeks post-injection, participants were categorised into a responder group (n = 108) and a non-responder group (n = 68). Data collect included demographic characteristics, botulinum toxin injection details, psychological counselling records, pre-injection assessment of cognitive function and self-efficacy. Pearson correlation analysis was used to assess the relationship between preoperative cognitive levels and self-efficacy, and the effectiveness of postoperative psychological counselling on treatment outcomes of anxiety and depression. Multivariate logistic regression analysis was employed to identify factors influencing treatment response to anxiety and depression, and receiver operating characteristic (ROC) curves were used to evaluate the predictive performance of these factors. Both study populations exhibited negative correlations between preoperative Pittsburgh Sleep Quality Index (PSQI) scores and Hamilton Anxiety Rating Scale (HAMA) reduction rates, as well as Hamilton Depression Rating Scale (HAMD) reduction rates. Conversely, preoperative Montreal Cognitive Assessment (MoCA) scores, preoperative self-efficacy, and duration per session showed positive correlations with HAMA reduction rates and HAMD reduction rates (all p < 0.05). Multivariate logistic regression analysis revealed that counselling frequency (OR = 3.808, β = 1.337), duration per session (OR = 1.092, β = 0.088), preoperative PSQI score (OR = 0.820, β = -0.198), MoCA (OR = 1.312, β = 0.272), and General SelfEfficacy Scale (GSES) score (OR = 1.175, β = 0.161) were identified as factors influencing treatment response of anxiety and depression following botulinum toxin injection (p < 0.05). ROC curve analysis indicated that the aforementioned variables possessed predictive value for treatment response. The combined predictive model yielded an area under the curve was 0.866 (95% confidence interval, ranging from 0.810 to 0.921). Preoperative cognitive function, selfefficacy and the duration per session were correlated with treatment response rates for anxiety and depression. Injection sites, counselling sessions, the duration per session, and preoperative PSQI, MoCA and GSES scores were identified as independent factors influencing treatment response following botulinum toxin injection.
Breast cancer (BC) patients shoulder considerable psychologicalstrain asthey traverse the illness trajectory. While anxiety and depression among this population have received substantial research attention, comprehensive global estimates that distinguish anxiety from depression remarkably scarce. Even more understudied is the systematic review of factors unique to each condition. Our meta-analysis was designed to redress these gaps by establishing worldwide pooled prevalence figures for both anxiety and depression among BC patients, alongside mapping their associated risk and protective influences. This study aimed to obtain consolidated global estimates of anxiety and depression prevalence within BC populations, and to explore various risk and protective factors that shape these. This systematic review and meta-analysis included cross-sectional and cohort studies from multiple databases reporting anxiety/depression prevalence in BC patients. Two investigators independently conducted study selection, data extraction, and quality assessment using the Newcastle-Ottawa Scale (NOS). A random-effects model pooled prevalence estimates; heterogeneity was explored via subgroup analysis and meta-regression. Publication bias was assessed with Egger's test and funnel plots. This meta-analysis included 32 studies comprising 21,507 BC patients. The pooled prevalence of anxiety was 35% (95% confidence intervals (CI): 30%-39%), and that of depression was 26% (95% CI: 23%-30%), with significant heterogeneity for both (p < 0.001). For anxiety, a high Life Orientation Test-Revised (LOT-R) score was protective, whereas low income was a risk factor. For depression, protective factors included older age, higher income, early tumor stage, and a high LOT-R score. Risk factors were low education, rural residence, disrupted marital status, comorbidities, lack of social support, and a history of mental illness. Sensitivity analysis confirmed that the results of this study were robust; although there was bias in anxiety and depression, its effect is limited after correction.  Conclusion: Anxiety and depression are highly prevalent in breast cancer patients, influenced by distinct sociodemographic and clinical factors, necessitating targeted psychological assessment and intervention.
Depressive disorders represent a major global health challenge, with inflammation and insulin resistance identified as key pathophysiological factors. The C-reactive protein-triglyceride glucose index (CTI), a novel composite biomarker integrating the inflammatory and metabolic pathways, has demonstrated enhanced predictive value in cardiometabolic diseases. However, its relationship with depression remains unexplored. This study examined the association between CTI and depressive symptoms in a nationally representative U.S. adult population. We conducted a cross-sectional analysis using National Health and Nutrition Examination Survey data from 2005 to 2023. Depressive symptoms were assessed using the Patient Health Questionnaire-9, with scores ≥10 indicating clinically significant symptoms. CTI was calculated as 0.412 × Ln(CRP) + Ln[triglycerides (mg/dL) × fasting glucose (mg/dL)/2]. Multivariable logistic regression models were employed to evaluate CTI-depression associations, adjusting for sociodemographic factors, comorbidities and laboratory parameters. Restricted cubic spline analysis assessed dose-response relationships, and subgroup analyses examined consistency across demographic and clinical strata. Among 15,318 participants (mean age 48.97 years; 49.78% female), 8.73% exhibited depressive symptoms. After comprehensive adjustment, each unit increase in CTI corresponded to a 23% increase in the risks of depression (odds ratio (OR) = 1.23, 95% confidence interval (CI): 1.11-1.36, p = 0.0001). Participants in the highest CTI tertile demonstrated 48% elevated odds compared with those in the lowest tertile (OR = 1.48, 95% CI: 1.17- 1.86, p = 0.0009), with a significant linear trend (p for trend = 0.0005). Restricted cubic spline analysis confirmed a linear dose-response relationship (p for nonlinearity = 0.1665). Associations remained consistent across age, sex, race/ethnicity and comorbidity subgroups (all p for interaction >0.05). Elevated CTI levels are independently associated with increased depression risk in U.S. adults, demonstrating a linear dose-response relationship. CTI may serve as a practical screening tool for identifying individuals at heightened depression risk, enabling integrated cardiometabolic-mental health interventions.
This study aimed to characterize depressive symptoms in patients with vascular cognitive impairment (VCI) after ischemic stroke and to identify independent predictors of treatment response to antidepressant therapy, with a focus on lesion-location heterogeneity. This retrospective observational cohort study enrolled 224 patients with VCI and concomitant depressive symptoms from June 2022 to June 2024. Depression severity was assessed using the 17-item Hamilton Depression Rating Scale (HAMD-17). All patients completed 8 weeks of standardized treatment including antidepressant medication and cognitive rehabilitation. Treatment response was defined as ≥50% reduction in HAMD-17 from baseline to week 8. Lesion locations were categorized into eight mutually exclusive anatomical groups (frontal, temporal, parietal, occipital, basal ganglia/internal capsule, thalamus, pons and cerebellum). Univariate and multivariable logistic regression identified predictors of treatment response, and Receiver Operating Characteristic (ROC) analysis evaluated model performance. Mean baseline HAMD-17 was 27.4 ± 3.1, and 65.2% (146/224) achieved treatment response. Baseline depressive symptom severity differed significantly across lesion locations (one-way analysis of variance (ANOVA): F (7,216) = 3.48, p = 0.001), whereas baseline anxiety severity did not (F (7,216) = 0.72, p = 0.652). In multivariable analysis, lower baseline Hamilton Anxiety Rating Scale (HAMA) score (odds ratio (OR) = 0.93, 95% confidence interval (CI): 0.88-0.98, p = 0.006), shorter time since stroke (OR = 0.86 per month, 95% CI: 0.75 to 0.99, p = 0.034), and higher education-adjusted Montreal Cognitive Assessment (MoCA) score (OR = 1.12 per point, 95% CI: 1.01-1.24, p = 0.031) were independently associated with treatment response. The prediction model demonstrated moderate discriminative ability (area under the curve (AUC) = 0.795, 95% CI: 0.738-0.851), with sensitivity of 0.890 and specificity of 0.608 at the optimal cutoff. Depressive symptom burden in poststroke VCI exhibits significant anatomical heterogeneity across lesion locations. Baseline anxiety severity, disease duration, and baseline cognitive performance moderately predict treatment response, supporting early risk stratification and individualized management.
To analyse the factors influencing anxiety and depression symptoms in patients with acute ischaemic stroke (AIS) and construct a nomogram model for predicting the risk of developing anxiety and depression. A total of 354 patients with AIS admitted to the Anhui Provincial Corps Hospital of Chinese People's Armed Police Force between May 2023 and April 2025 were selected. They were divided into anxiety and depression (n = 180) and nonanxiety and depression (n = 174) groups on the basis of the presence of anxiety (Hamilton Anxiety Rating Scale ≥ 7 points) and depression (Hamilton Depression Rating Scale ≥ 8 points). Baseline patient data, laboratory indicators and anxiety-depression status were extracted from medical records. Logistic multivariable analysis was employed to identify factors influencing the occurrence of anxiety and depression in patients with AIS, establishing a nomogram prediction model. Internal validation was conducted by using receiver operating characteristic (ROC) and calibration curves and decision curve analysis. Comparative analysis revealed statistically significant differences between the two patient groups in terms of age, neutrophil count, angiogenin-like protein 4 (ANGPTL4), silencing information regulator protein 1 (SIRT1), Krüppel-like transcription factor 2 (KLF2), retinol-binding protein (RBP), lipoprotein (a) levels and plaque stability (p < 0.05). Multivariate logistic regression analysis revealed that age (odds ratio [OR] = 1.311, 95% confidence interval [95%CI]: 1.031-1.667), ANGPTL4 (OR = 0.057, 95%CI: 0.023-0.144), SIRT1 (OR = 0.096, 95%CI: 0.016-0.554), KLF2 (OR = 0.001, 95%CI: 0.000- 0.401), RBP (OR = 1.476, 95%CI: 1.068-2.040), lipoprotein (a) (OR = 1.130, 95%CI: 1.024-1.247) and plaque stability (OR = 23.941, 95%CI: 5.178-32.186) were factors influencing anxiety and depression symptoms in patients with AIS (p < 0.05). The established regression model equation is log(P) = 0.271 × age - 0.044 × ANGPTL4 - 2.348 × SIRT1 - 7.453 × KLF2 + 0.390 × RBP + 0.122 × lipoprotein (a) + 3.176 × plaque stability. The area under the ROC curve was 0.901 (95%CI: 0.867-0.934). Internal validation using the bootstrap method demonstrated high concordance between the predictive and standard model curves. Age, ANGPTL4, SIRT1, KLF2, RBP, lipoprotein (a) and plaque stability are factors influencing the occurrence of anxiety and depression symptoms in patients with AIS. The nomogram model constructed on the basis of these factors demonstrates predictive validity.
Social anxiety is a key risk factor for adolescent depression, yet its underlying mechanisms and subgroup differences remain unclear. This study explored the mediating roles of perceived social support and sleep quality in their link, and the moderating effects of visit type and gender. A retrospective observational study enrolled 386 depressed adolescents (12-18 years; 231 outpatients, 155 inpatients) from Xiamen Xianyue Hospital. Social anxiety, depressive symptoms, perceived social support and sleep quality were assessed using the Social Anxiety Scale for Adolescents, Self-Rating Depression Scale, Multidimensional Scale of Perceived Social Support and Pittsburgh Sleep Quality Index, respectively. Pearson's correlation and Hayes' PROCESS macro (Model 6) were applied for mediation/moderation analyses, with sensitivity testing via the Montgomery-Åsberg Depression Rating Scale (MADRS). Inpatients and severe cases had higher Social Anxiety Scale for Adolescents (SAS-A), Zung Self-Rating Depression Scale (SDS) and Pittsburgh Sleep Quality Index (PSQI) and lower Multidimensional Scale of Perceived Social Support (MSPSS) scores (all p < 0.001); females had higher social anxiety (p = 0.003). Social anxiety correlated positively with depressive symptoms (r = 0.54, p < 0.001) and negatively with perceived social support (r = -0.49, p < 0.001). Mediation analysis showed a total effect of social anxiety on depressive symptoms (β = 0.304, p < 0.001), with direct effect (57.5%, β = 0.175) and total indirect effect (42.5%, β = 0.130). Key indirect pathways: 'social anxiety → sleep quality → depressive symptoms' (27.0%) and a serial mediation pathway via perceived social support and sleep quality (8.2%); perceived social support's single mediation was marginally non-significant (p = 0.118). Moderation analyses revealed stronger direct effects in inpatients (β = 0.342 vs. 0.097, p < 0.001) and stronger sleep quality effects in females (β = 1.125 vs. 0.619, p = 0.006). MADRS sensitivity analyses confirmed consistency (path coefficient deviations <1%). Social anxiety affects adolescent depressive symptoms directly and via sleep-related mediation, moderated by visit type and gender. Targeting social anxiety and sleep quality may optimise precision prevention/treatment for adolescent depression.
Suicidal ideation and non-suicidal selfinjury (NSSI) are highly prevalent during early adolescence and represent serious public health concerns. Emotional dysregulation has been identified as a transdiagnostic factor underlying various risk behaviors, particularly in youth with externalizing difficulties such as attentiondeficit/hyperactivity disorder (ADHD). While prior research has examined these factors independently, few studies have explored their joint and differential contribution to suicide-related outcomes in early adolescence. This study aimed to examine the association of ADHD-related symptoms, positive and negative affect, and emotional dysregulation with suicidal ideation and NSSI, as well as to explore whether emotional dysregulation moderates the relationship between ADHD-related symptoms and these risk behaviors. A total of 1079 Spanish adolescents (Mage = 12.6 years, SD = 0.6) enrolled in the first year of compulsory secondary education participated in a cross-sectional study. Standardized self-report measures were used to assess ADHD-related symptoms (Strengths and Difficulties Questionnaire [SDQ]), affect (Positive and Negative Affect Schedule for Children and Adolescents [PANAS-N]), and difficulties in emotion regulation (Difficulties in Emotion Regulation Scale [DERS-18]), alongside indicators of suicidal ideation and NSSI. Data analysis included group comparisons and hierarchical logistic regression models. Suicidal ideation was reported by 10.1% of participants, and NSSI by 16.4%. Adolescents with either behavior exhibited significantly more hyperactivityinattention symptoms, greater negative affect, lower positive affect, and substantially elevated emotional dysregulation. In regression analyses, ADHD-related symptoms remained significant predictors after accounting for affect, although emotional dysregulation emerged as the strongest predictor, reducing the effects of other variables. Specific dimensions such as lack of emotional clarity and nonacceptance of emotional responses were associated with suicidal ideation, while impulse control difficulties and lack of adaptive strategies predicted NSSI. No moderation effects were found. ADHD-related symptoms and emotional dysregulation both contribute independently to adolescent suicidal ideation and NSSI, with emotional dysregulation showing the most robust predictive value. Schoolbased prevention efforts should incorporate emotion regulation skill-building to reduce suicide risk and NSSI in early adolescence.
To develop machine learning-based prediction models for postoperative depression risk in patients with ovarian cancer and to evaluate their predictive performance and clinical application value. To develop machine learning-based prediction models for postoperative depression risk in patients with ovarian cancer and to evaluate their predictive performance and clinical application value. Among 850 patients, 268 (31.5%) were positive for postoperative depression risk. Feature selection identified 13 predictive variables: age, operation time, length of hospital stay, pain score, white blood cell count, albumin, C-reactive protein, CA125, education level, history of depression/anxiety, postoperative insomnia, fatigue, and opioid analgesic use. Among the five models, random forest demonstrated superior performance with an AUC of 0.776 in the validation set, a Brier score of 0.182, sensitivity of 0.771, and an F1 score of 0.792, along with satisfactory calibration and clinical net benefit. SHAP analysis revealed that pain score, postoperative insomnia, albumin level, and opioid use contributed substantially to model predictions. A nomogram based on logistic regression model was constructed for intuitive individual risk assessment. The machine learning-based prediction models for postoperative depression risk in patients with ovarian cancer demonstrated satisfactory discriminative ability and clinical utility, with random forest model showing optimal performance. A clinical nomogram was additionally constructed to enable individualised and visual risk quantification suitable for bedside application. Together, these tools facilitate early identification of high-risk patients and provide evidence for clinical intervention.
This study aimed to analyse the effect of sleep quality on cognitive function in elderly patients with Alzheimer's disease (AD). This retrospective study extracted clinical data from the hospital's electronic medical record system for elderly patients with AD admitted to the Departments of Neurology or Geriatrics between June 2022 and June 2024. Cognitive function was assessed using the MiniMental State Examination (MMSE), subjective sleep quality was evaluated with the Athens Insomnia Scale (AIS) and objective sleep architecture parameters were measured via overnight polysomnography (PSG). Participants were stratified into mild and moderate-to-severe cognitive impairment groups according to their MMSE scores. General characteristics and sleep-related indicators were compared between the two groups. A binary logistic regression model was employed to analyse independent factors influencing cognitive impairment severity. In this model, cognitive impairment severity served as the dependent variable, and PSG parameters and AIS score served as the core independent variables. Adjustments were made for potential confounding factors, including age, gender, years of education, disease duration, Hospital Anxiety and Depression Scale (HADS) scores and  Instrumental Activities of Daily Living Scale (IADL) scores. The cohort comprised 61 (40.67%) moderate-to-severe and 89 (59.33%) mild impairment patients. Compared with the mild impairment group, the moderate-to-severe group showed significantly poorer subjective (higher AIS) and objective sleep profiles, including reduced total sleep time, efficiency, and N2/N3 sleep and increased N1 sleep, latency and awakenings (p < 0.05). Adjusted regression identified the N3 stage/total sleep time ratio as a protective factor (odds ratio [OR] = 0.720, 95% CI: 0.576-0.900, p = 0.004) and the AIS score (OR = 1.850, 95% CI: 1.405-2.434, p < 0.001) and number of awakenings (OR = 3.101, 95% CI: 1.879-5.116, p < 0.001) as independent risk factors. In elderly patients with AD, impaired objective sleep architecture and subjective insomnia are significantly associated with poor cognitive function. This study highlighted sleep parameters as potential indicators for cognitive status assessment.
Postherpetic neuralgia (PHN) remains the most frequent and distressing sequela of herpes zoster (HZ). Although psychological factors are known to contribute to chronic pain, their role during the acute phase of HZ in predicting PHN has not been fully clarified. This study aimed to investigate whether acute-phase anxiety is independently associated with subsequent PHN development and to evaluate the added predictive value of anxiety assessment in clinical prediction models. This longitudinal cohort study enrolled 99 patients with acute HZ between January 2022 and January 2025. Anxiety was assessed at baseline using the Hospital Anxiety and Depression Scale-Anxiety subscale (HADS-A) and the State-Trait Anxiety Inventory-State subscale (STAI-S). PHN was defined as pain lasting at least 90 days after rash onset. We performed logistic regression to identify independent predictors and constructed receiver operating characteristic (ROC) curves to evaluate predictive performance. The 3 months incidence of PHN was 41.4% (41/99). Patients who developed PHN had significantly higher baseline HADS-A scores (10.7 ± 3.8 vs. 6.4 ± 3.2, p < 0.001) and STAI-S scores (52.8 ± 10.4 vs. 41.3 ± 9.7, p < 0.001) than those who did not. Multivariate analysis identified three independent predictors: age (adjusted odds ratio [OR] = 1.47 per 10 years, p = 0.038), baseline pain intensity (adjusted OR = 1.41, p = 0.012), and HADS-A score (adjusted OR = 1.31 per unit, p = 0.001). Clinically significant anxiety (HADS-A ≥ 8) was associated with a 4.52-fold increased risk of PHN (95% confidence interval: 1.72-8.58, p = 0.013). HADS-A demonstrated good discriminative ability (area under the curve [AUC] = 0.813), and a combined model incorporating anxiety with clinical variables achieved superior predictive performance (AUC = 0.876 vs. 0.762, p = 0.019). Acute-phase anxiety is an independent and clinically meaningful predictor of PHN. Adding anxiety assessment to clinical evaluation significantly improves risk prediction accuracy, supporting routine psychological screening and early intervention in patients with HZ.
Terminal colon adenocarcinoma is a debilitating condition often accompanied by severe pain and substantial anxiety and depression. Hospice care provides a dedicated framework to address this symptom complex, yet robust evidence for its real-world effectiveness within the Chinese healthcare context remains underdeveloped and insufficiently documented. This study aimed to evaluate the effects of hospice care on key clinical outcomes in terminal colon adenocarcinoma. This retrospective cohort analysis reviewed data from 92 patients with histologically confirmed terminal colon adenocarcinoma (≥stage Ⅲ) treated at The Affiliated Yangming Hospital of Ningbo University between January 2024 and June 2025. The cohort included 46 patients receiving integrated hospice care alongside standard oncology treatment and 46 matched controls receiving standard care only. Comparative analyses of depression (Hospital Anxiety and Depression Scale-Depression, HADS-D), anxiety (HADS-A), pain (visual analogue scale, VAS), opioid usage, and healthcare utilisation were conducted at baseline, 1, 3, and 6 months. Baseline characteristics were comparable between groups. HADS-D scores decreased more in the hospice care group (from 9.13 ± 3.39 to 5.91 ± 2.72) than in the standard group (from 9.35 ± 3.58 to 8.57 ± 3.04; p < 0.001). HADS-A scores showed a greater reduction in the hospice care group (from 8.63 ± 2.55 to 5.83 ± 2.57) than in the standard group (from 9.13 ± 3.17 to 8.48 ± 2.63; p < 0.001). The hospice care group demonstrated significantly greater reductions in VAS scores (from 6.83 ± 1.19 to 3.17 ± 1.01) compared with the standard group (from 6.74 ± 1.40 to 5.51 ± 1.63; p < 0.001) and a higher proportion achieved ≥30% pain reduction at 6 months(80.43% vs. 39.13%, p < 0.001). Additionally, hospice care was associated with lower opioid consumption, shorter hospital stays, fewer emergency visits, and reduced re-admissions (all p < 0.05), with no increase in adverse events. For patients with terminal colon adenocarcinoma, integrated hospice care was associated with significantly improved pain control and reduced anxiety and depressive symptoms. It was also associated with decreased healthcare utilisation with a favourable safety profile.
To examine the Tanner stage-specific patterns of psychological and behavioural problems in girls with central precocious puberty (CPP) and to identify clinical predictors of clinically significant psychological impairment. This retrospective cross-sectional study included 116 girls with CPP treated between January 2020 and December 2024. Clinical data were extracted from electronic medical records. Pubertal development was classified by breast Tanner staging into stage II (n = 32), stage III (n = 42) and stages IV-V (n = 42). Psychological outcomes were evaluated using the age-standardised Child Behavior Checklist (CBCL) T-scores. For participants aged ≥7 years, depressive and anxiety symptoms were assessed using the Children's Depression Inventory and the Screen for Child Anxiety Related Emotional Disorders. The independent predictors of clinically significant psychological problems were identified through logistic regression analysis, defined as CBCL Total Problems T-score of ≥64, and model performance was assessed using receiver operating characteristic curve analysis. The mean CBCL Total Problems T-score was 58.6 ± 12.4, and 50 girls (43.1%) had clinically significant psychological problems. Psychological burden increased across Tanner stages. The CBCL Total T-scores were 52.4 ± 10.8 in stage II, 58.6 ± 11.6 in stage III and 65.8 ± 13.2 in stages IV-V (p < 0.001). The prevalence of clinically significant problems increased from 21.9% to 40.5% and 61.9% (p = 0.002). Internalising symptoms showed a stronger stage-related pattern than externalising symptoms. Early age at onset was associated with increased symptom severity (r = -0.385, p < 0.001). Multivariable analysis identified Tanner stages IV-V versus stage II (odds ratio [OR] = 2.94, 95% confidence interval [CI]: 1.44-5.99, p = 0.003), younger age at onset (OR = 0.72, 95% CI: 0.55-0.94, p = 0.016), higher peak luteinising hormone (LH) (OR = 1.45, 95% CI: 1.08-1.95, p = 0.014) and family history of early puberty (OR = 2.18, 95% CI: 1.26-3.77, p = 0.005) as independent predictors. The prediction model showed good discrimination (area under the curve = 0.812, 95% CI: 0.738-0.886). Girls with CPP experience substantial psychological burden, and risk increases with pubertal stage. Advanced Tanner stage, earlier pubertal onset, increase in peak LH and family history are indicators for enhanced psychological monitoring and support.
Elderly individuals frequently suffer from chronic underlying disease like hypertension and diabetes. These illnesses not only impair physical health but also show a close link to elevated rates of cognitive impairment and depressive symptoms. The mutual influence between these two issues can further diminish the quality of life of elderly patients. Nevertheless, there remains a shortage of systematic investigations into the cognitiveemotional relationship within this specific population. A total of 206 elderly patients with chronic underlying disease were enrolled retrospectively. Cognitive function was assessed using the Montreal Cognitive Assessment (MoCA). Depressive symptoms were evaluated via the 15-item Geriatric Depression Scale (GDS-15). To analyse the correlation between cognitive impairment and depressive symptoms, the Spearman correlation analysis, univariate analysis, and multivariate Logistic regression were employed. Among the 206 patients, the primary chronic underlying diseases were hypertension (64.56%), diabetes mellitus (39.32%), coronary heart disease (27.67%), and chronic obstructive pulmonary disease (21.84%). Additionally, 45.15% of the patients had two or more chronic diseases. The prevalence of cognitive impairment stood at 49.51% (102 cases), while the prevalence of depressive symptoms was 40.78% (84 cases). No significant differences were observed in the prevalence of cognitive impairment and depressive symptoms among patients with different types of chronic diseases (all p > 0.05). Spearman correlation analysis revealed a significant negative correlation between MoCA scores and GDS-15 scores (r = -0.552, p < 0.001). Binary Logistic regression analysis indicated that factors such as body mass index ≥21.42 kg/m2 , number of chronic diseases ≥2.5, GDS-15 score ≥4.5 points, and Multidimensional Scale of Perceived Social Support score ≥52.5 were independent risk factorsfor moderate-to-severe cognitive impairment (all p < 0.05).  Conclusions: The severity of cognitive impairment in elderly patients with comorbid chronic underlying diseases increases with the exacerbation of depressive symptoms.
Antidepressant use has been associated with adverse skeletal outcomes in observational studies, but whether this association is causal remains unclear due to potential confounding factors. This study employed Mendelian randomisation (MR) to systematically evaluate the causal associations between genetically predicted antidepressant use and the risks of osteoporosis and fractures at multiple anatomical sites. Using publicly available genome-wide association studies (GWAS) datasets, we defined genetically predicted antidepressant use as the exposure and osteoporosis and fractures (spine, leg, and wrist) as outcomes. Genome-wide significant single nucleotide polymorphisms (SNPs) were selected as instrumental variables following dataset harmonisation. The inverse-variance weighted (IVW) method was used as the primary MR approach, supplemented by MR-Egger regression, weighted median, and weighted/simple mode methods. Heterogeneity tests, funnel plots, and leave-one-out sensitivity analyses were performed to assess the robustness and consistency of the results. MR analysis demonstrated significant positive causal associations between antidepressant use and osteoporosis across two independent datasets (ukb-a-87 and ukb-b-12141), with IVW odds ratios of 1.0035 and 1.0016, respectively. Genetically proxied antidepressant use showed significant causal effects on fractures at all examined sites, with wrist fractures displaying the strongest association (IVW: odds ratio (OR) = 1.0027, p = 0.0041). Effect directions remained consistent across multiple MR methods, with no significant heterogeneity detected. Leave-one-out analyses confirmed no single SNP disproportionately influenced the results. This MR study provides evidence that antidepressant use may directly influence bone metabolism and increase fracture susceptibility, particularly at the wrist. These findings highlight the importance of bone health monitoring in patients receiving antidepressant therapy, especially those at elevated fracture risk. Further mechanistic studies and longitudinal validation are warranted.