PurposeThe impact of the COVID-19 pandemic on young adults with serious mental illnesses is poorly understood despite the high likelihood that important milestones for this group were affected by the pandemic and its economic aftermath. This study explored how U.S. young adults with disability due to mental health disorders experienced transitions to adulthood before and during the pandemic.MethodsSemi-structured qualitative interviews with 20 emerging adults, age 18-25 (mean age, 22.3) including 10 female, 4 male, and 6 gender non-binary or other gender individuals, covered 7 life transitions identified from the literature: residential independence, school completion, friendship network development, individuation from family, full-time employment, marriage/intimate partner establishment, and parenthood. Summative content analysis and inductive thematic analysis were conducted by a team including young adult co-researchers with lived experience of mental health challenges and disabilities.ResultsA substantial number of transitions completed before the pandemic included graduation, residential independence, forming friendship networks, and individuation from family. Lockdown often derailed transitions, especially education and employment; and led to abandonment or postponement of other transitions, especially romantic relationships and full-time work. Educational interruptions reverberated beyond academics to include food insecurity, economic instability, and residential uncertainty. Life course concerns included not meeting cultural expectations of timely transitions, and distress over the disconnect between personal aspirations and current circumstances. Resilience was also observed, especially concerning mutual assistance from social networks and feelings of pride at being able to successfully negotiate pandemic stressors.ConclusionDespite significant psychiatric multi-morbidity and pandemic-related disruptions, participants demonstrated a largely normative pre-pandemic life trajectory and showed notable resilience in the face of adversity. Findings suggest the need for transition-focused wraparound service systems, early intervention strategies to avoid development of further comorbidities, and supports incorporating informal social networks such as friends and family.
ObjectivePositive youth development (PYD) can be enhanced through dynamic interactions between individuals and their environments. While previous research has primarily focused on its social-emotional learning mechanisms, the cognitive and motivational pathways of PYD were seldom explored. To address this gap, this study investigated the associations of parent -child relationships, teacher–student relationships, and friendships with adolescents' prosocial behavior, with a focus on the mediating role of meaning in life.MethodsA three-wave temporally ordered mediation analysis was conducted using three waves of data collected at six-month intervals from 671 Chinese junior high school students.ResultsThe findings revealed that the quality of parent–child relationships and friendships at Wave 1 was positively associated with both the presence and search for meaning in life at Wave 2, with friendships showing stronger associations, whereas teacher–student relationships were not significantly associated with either dimension of meaning in life. Furthermore, only the presence of meaning in life at Wave 2 was positively associated with prosocial behavior at Wave 3, while the search for meaning in life was not significantly associated with prosocial behavior.ConclusionThese findings provide practical insights into China's moral education framework, rooted in Confucian doctrine, by suggesting that high-quality relationships are associated with adolescents' sense of societal contribution as an aspect of life's meaning. Possible reasons for the nonsignificant roles of teacher–student relationships and the search for meaning in life were discussed, along with the study's limitations and directions for future research.
BackgroundNeuroimaging studies of internet gaming disorder (IGD) have primarily focused on reward-related and executive-control networks. However, sensory processing of addiction-related cues, particularly visual processing, may also contribute to IGD. Whether the intrinsic hierarchical organization of the visual system is altered remains unclear.MethodsThe study included 35 individuals with IGD and 30 healthy controls (HCs), with an overall age range of 9–33 years. The sample was predominantly composed of adolescents. An a priori three-level visual hierarchy model comprising the primary visual cortex (V1), lingual gyrus, and lateral occipital cortex (LOC) was constructed. Functional connectivity (FC) was calculated for the V1–Lingual, Lingual–LOC, and V1–LOC pathways. A hierarchical integration index (HII), derived from V1–LOC and V1–Lingual connectivity, was calculated to assess long-range relative to local visual coupling. Group differences were assessed using MANCOVA with follow-up univariate analyses and FDR correction. Within the IGD group, associations between IAT scores and all visual hierarchical measures were examined using Pearson and covariate-adjusted partial correlations with FDR correction.ResultsV1–LOC connectivity was significantly reduced in the IGD group compared with HCs (HC: 0.364 ± 0.132; IGD: 0.130 ± 0.189; F(1,59) = 30.809; FDR-corrected p < 0.001; partial η2 = 0.343). V1–Lingual and Lingual–LOC connectivity did not show significant group differences, although their effect sizes were small-to-moderate (Cohen’s d = 0.32 and 0.40, respectively). The HII was numerically lower in IGD but did not reach statistical significance (t = 1.90, p = 0.062; Cohen’s d = 0.46). Within the IGD group, lower V1–LOC connectivity was associated with higher IAT scores (r = −0.479, FDR-corrected p = 0.016), and this association remained significant after controlling for age, sex, education, and mean FD (partial r = −0.447, FDR-corrected p = 0.048).ConclusionsIGD is associated with altered long-range hierarchical connectivity within the visual system, particularly between early visual cortex and higher-order LOC. These findings provide a sensory-level perspective complementing reward- and control-centered accounts of IGD.
BackgroundNon-suicidal self-injury (NSSI) is clinically important among people with mood disorders. We developed an interpretable machine-learning model for NSSI during one-year follow-up and evaluated its internal robustness.MethodsThe cohort comprised 912 inpatients with mood disorders (MD) and 72 candidate predictors. We collected demographic information, disease characteristics and laboratory indicators of MD patients who were hospitalized throughout 2022, and followed them up one year after discharge to collect their NSSI incidence within one year. Least Absolute Shrinkage and Selection Operator (LASSO) regression was utilized to identify predictors associated with NSSI features. Subsequently, multiple machine learning algorithms were constructed, and the optimal model was selected based on its area under the curve (AUC), F1-score, and accuracy. Finally, Shapley Additive Explanations (SHAP) were applied to the optimal model to rank feature importance and interpret the directional impact of each selected predictor on NSSI risk.ResultsFifty-two participants (5.70%) experienced NSSI. Across 25 repeated internal evaluations, random forest (RF) had the highest mean AUPRC (0.456; SD 0.116) and the most fold-level AUPRC wins (8/25) and was therefore designated the final model. The prediction of RF achieved a ROC-AUC of 0.861 (95% CI 0.738-0.951), AUPRC of 0.285 (95% CI 0.172-0.535) and Brier score of 0.070; at the locked threshold, recall was 0.091 and F1 was 0.133. RF SHAP importance was highest for age, previous NSSI, FT4, current NSSI and estradiol.ConclusionRF showed the strongest repeated AUPRC among the evaluated algorithms and was selected as the final model, but the small number of events, very limited case detection at the locked threshold and absence of external validation preclude clinical deployment.
BackgroundGeneral-purpose large language models are increasingly used by patients and caregivers to obtain mental health information and guidance about when professional care is required. In late-life depression, broadly accurate information may nevertheless be unsafe when cognitive change, multimorbidity, frailty, polypharmacy, self-neglect, caregiver dependence, or suicide risk is not adequately recognized. This study compared the clinical accuracy, safety, geriatric-specific appropriateness, triage performance, and response consistency of three large language models when answering patient- and caregiver-centered questions about late-life depression.MethodsWe conducted a blinded, paired benchmarking study using 90 questions covering six geriatric psychiatry domains and equally distributed across low-, moderate-, and high-risk strata. Each question was submitted independently to GPT-5.5 Instant via ChatGPT, Gemini 3.5 Flash via Gemini, and Seed2.0 Pro via Doubao, generating 270 primary-round responses. A stratified subset of 30 questions was resubmitted in separate conversations to assess test-retest consistency, yielding 360 responses overall. Two psychiatrists independently evaluated anonymized outputs against prespecified item-specific reference standards, with clinically important disagreements adjudicated by a third senior psychiatrist. The primary outcome was the proportion of clinically acceptable responses, defined using accuracy, clinical safety, geriatric appropriateness, warning-sign recognition, and triage criteria.ResultsClinically acceptable responses were generated for 78.9% of questions by ChatGPT, 72.2% by Gemini, and 60.0% by Doubao (overall P<0.001). The difference between ChatGPT and Gemini was not statistically significant, whereas both outperformed Doubao. This model ranking remained unchanged under alternative core-safety and more stringent optimal-response definitions. Complete geriatric-specific appropriateness was achieved in 64.4%, 55.6%, and 43.3% of responses, respectively. Major safety errors occurred in 5.6% of ChatGPT responses, 10.0% of Gemini responses, and 16.7% of Doubao responses (raw P = 0.015; FDR-adjusted q=0.023). Performance declined substantially with increasing clinical risk. In exploratory post hoc caregiver-centered analyses, clinical acceptability was 76.7% for ChatGPT, 70.0% for Gemini, and 53.3% for Doubao, while explicit caregiver-directed action was present in 80.0%, 70.0%, and 53.3% of responses, respectively. Among high-risk questions, clinically acceptable response rates were 63.3%, 50.0%, and 36.7%, while under-triage occurred in 16.7%, 26.7%, and 36.7% of responses, respectively. Test-retest clinical consistency was highest for ChatGPT (90.0%), followed by Gemini (83.3%) and Doubao (73.3%).ConclusionsThe three models answered many late-life depression questions accurately and safely, but none demonstrated consistently reliable performance across complex or high-risk scenarios. Clinically important weaknesses involved geriatric-specific interpretation, recognition of indirect risk signals, crisis-response completeness, under-triage, and response stability. General-purpose large language models may support selected low-risk educational tasks but should not independently guide emergency triage, medication changes, suicide-risk management, or other safety-critical decisions in older adults.
