Cognitive reappraisal deficits are a transdiagnostic risk factor for major depressive disorder (MDD) and social anxiety disorder (SAD) and are observed in patients with these disorders at the neural level. Preliminary research suggests less activation of prefrontal regions during reappraisal (vs. viewing) of negative stimuli associates with overall symptom severity in patients with MDD or SAD, however, this is not reliably observed across studies. Consistent with research showing that reappraisal may only be adaptive when employed to cope with uncontrollable adversity, this study sought to examine whether neighborhood-level adversity (i.e. socioeconomic disadvantage, crime) moderated the relation between internalizing symptom severity and neural correlates of reappraisal.

This study included patients with a current diagnosis of MDD (n = 51) or SAD (n = 39). Patients completed measures of symptom severity as well as an emotion regulation task while in the scanner to assess neural activation during reappraisal. Patients’ addresses were geocoded to assess neighborhood socioeconomic disadvantage and crime.

Results indicated that greater symptom severity was associated with decreased activation of key prefrontal regions underlying reappraisal, but only for patients living in neighborhoods characterized by high levels of personal (i.e. violent) crime. Unexpectedly, the opposite was found for patients living in low-crime neighborhoods, such that greater symptom severity was associated with increased neural activation during reappraisal (vs. viewing) of negative stimuli.

Findings highlight the critical importance of considering patients’ neighborhood contexts when evaluating associations between symptom severity and neural correlates of reappraisal in patients with internalizing disorders.

This study evaluated the effectiveness of a one-year smoking cessation intervention for people with severe mental illness compared with treatment as usual (TAU) in outpatient mental healthcare.

The KISMET study is a pragmatic cluster-randomized controlled trial conducted in 21 outpatient mental healthcare teams in the Netherlands. Eleven teams delivered the KISMET intervention comprising cognitive-behavioral and peer support, combined with optional pharmacological reatment. Ten teams participated in the TAU condition. We collected data between October 2022 and July 2024. The primary outcome was self-reported smoking cessation at 12 months, verified through exhaled carbon monoxide levels below 10 parts per million. Secondary outcomes included depression and anxiety (HADS), severity of psychotic symptoms (PANSS-6), quality of life (SF-12), disease self-management (PAM-13), lipid profile, blood pressure, body mass index, glucose level, and physical fitness. Crude and adjusted linear and multivariable logistic regression and mixed model analyses were performed.

Eighty-nine participants were included in the KISMET intervention and 44 in TAU. Smoking cessation rates were significantly higher in the KISMET group at 3 months (OR 12.1, 95% CI 1.4 to 103.7) and at 12 months (OR 4.2, 95% CI 1.0 to 17.2) but not at 6 months (OR 1.9, 95% CI 0.5 to 6.9). No significant differences between groups were found for secondary outcomes. Dropout rates were 58% in the intervention and 32% in the TAU group.

The KISMET intervention shows potential without signs of physical or psychopathological complications. However, results must be interpreted with the high dropout rates in mind.

The COVID-19 pandemic raised concerns about the mental health of an already burdened healthcare workforce. This study examined mental health trajectories among healthcare workers (HCWs) across the pandemic and identified personal and employment factors associated with different symptom patterns.

Longitudinal data were drawn from the NHS CHECK cohort, including clinical and non-clinical staff from 18 NHS Trusts in England (April 2020–April 2023). Growth curve and growth mixture models identified latent classes of HCWs characterized by distinct trajectories of probable common mental disorders. Secondary outcomes included anxiety, depression, alcohol misuse, and post-traumatic stress symptoms. Logistic regression examined associations between baseline personal and employment characteristics and class membership.

The analytical sample included 22,764 participants. For each outcome, growth mixture models identified two latent classes. Approximately 31% of HCWs experienced persistently high symptoms of probable common mental disorders, while 69% experienced persistently low symptoms. Similar patterns were observed for secondary outcomes, with small subgroups demonstrating worsening symptoms followed by improvement. Logistic regression analyses showed that being female, younger, single, working as a nurse, or having a pre-existing mental health diagnosis increased the odds of belonging to a high symptom class. Perceived support from colleagues and managers was protective.

While many HCWs reported consistently low mental health symptom levels, almost a third belonged to a latent class characterized by persistently high symptoms across all time points. These findings underscore the need for mental health support for vulnerable HCW groups, embedded within routine NHS practice rather than limited to crisis periods.

Recent proposals for revising the Diagnostic and Statistical Manual of Mental Disorders (DSM) aim to improve psychiatric diagnosis. While these efforts reflect substantial ambition, they continue to operate within assumptions embedded in the DSM’s underlying classificatory logic. This editorial examines whether such incremental revision is sufficient.

We provide a critical analysis of the recently published DSM roadmap and accompanying subcommittee commentaries. Drawing on contemporary literature, we identify five structural blind spots in the current reform agenda: public mental health, scientific inference, lived experience, epistemic governance, and the function of diagnosis. Based on this analysis, we propose an alternative dialogical redesign for the DSM.

We argue that current revision considerations risk increasing complexity without resolving fundamental limitations in psychiatric classification. Specifically, our analysis highlights several areas that warrant further consideration, including the relationship between diagnostic expansion and societal conditions, the applicability of group-level scientific findings to individual care, the incorporation of experiential knowledge, participatory governance in revision processes, and the identity-related implications of diagnosis. In response, we propose redesigning the DSM as a hybrid dialogical system that retains coarse-grained classificatory categories for pragmatic purposes while shifting diagnostic practice toward contextual interpretation, collaborative meaning-making, relational understanding, and individualized care formulation.

The challenges facing psychiatric diagnosis require more than incremental refinement. We therefore argue for a dialogical redesign of the DSM that better reflects the context-dependent, experiential, and relational nature of mental health conditions, positioning diagnosis as a starting point for collaborative inquiry.

Early prediction of depressive and anxiety disorders is challenging due to substantial heterogeneity in risk pathways. Conventional machine-learning models trained on aggregated populations may obscure subgroup-specific mechanisms and limit interpretability for prevention. We evaluated whether a hybrid unsupervised–supervised framework can identify meaningful subgroups and yield more interpretable risk prediction.

We analyzed cohort data of 15,897 Japanese adults who completed baseline (August–September 2020) and 6-month follow-up (February–March 2021) surveys and did not screen positive for depressive and anxiety disorders at baseline (K6 score < 13). Using 169 baseline demographic, psychosocial, lifestyle, and behavioral variables, we performed hierarchical clustering to derive data-driven subgroups. Within each cluster, we trained Random Forest models to predict incident screened depressive and anxiety disorders at follow-up (K6 ≥ 13) and interpreted predictors using SHapley Additive exPlanations (SHAP).

The overall 6-month incidence was 6.23%. A five-cluster solution revealed two high-risk subgroups: an older-adult profile with poor quality of life (12.9%) and a working-parent profile characterized by work–family overload (29.8%). Compared with a global model trained on the full sample, the cluster-then-predict framework showed broadly similar overall performance but performed better in the highest-risk subgroup and revealed more differentiated predictor profiles. Loneliness, health-related quality of life, happiness, and personality traits predominated in clusters with moderate adversity, whereas lifestyle disruption (sleep, diet, and irregular routines) characterized the high-risk late-life subgroup and alcohol dependence and work–family burden characterized the high-risk working-parent subgroup.

Addressing risk-factor heterogeneity before prediction may enable more interpretable, context-tailored prevention strategies.

Schizophrenia (SCZ) and bipolar disorder (BD) share substantial clinical and neuroanatomical features, yet the neurogenetic basis underlying their shared gray matter volume (GMV) deficits remains poorly understood.

We conducted meta-analyses to identify convergent GMV alterations across the two disorders. Genome-wide association studies (GWAS) were performed to uncover genetic variants associated with the shared GMV deficits region in UK Biobank participants. Polygenic risk score (PRS)-GMV associations were analyzed to examine the cumulative influence of genetic risk on GMV in regions with shared deficits. Furthermore, pleiotropic SNPs jointly associated with SCZ, BD, and shared GMV deficits were identified. Spatiotemporal gene expression profiling was utilized to characterize the developmental trajectories, and molecular docking was performed to explore potential drugs.

Meta-analysis revealed consistent overlapping GMV reductions in frontal, temporal, and insular regions across SCZ and BD, based on 6,620 patients and 7,762 controls. GWAS identified 14 SNPs associated with the shared GMV deficits. PRS analyses showed that modestly higher SCZ polygenic risk correlated with decreased GMV of shared regions. Two pleiotropic SNPs – rs11191368 and rs79668541 – were linked to both disorders and the shared GMV deficits. Spatiotemporal expression analyses demonstrated distinct developmental trajectories, and molecular docking highlighted 168 drugs with binding interactions for shared genes.

This study delineates shared neurogenetic mechanisms linking GMV abnormalities to genetic risk across SCZ and BD. Given the cross-sectional design, future longitudinal studies in independent cohorts are warranted to validate these findings and clarify the temporal relationships.

Recent research indicates that neuropathological alterations may propagate via brain networks, as illustrated by network diffusion models (NDMs). The application of NDM to internet gaming disorder (IGD) has yet to be evaluated. This study was set to identify possible epicenters of neuroanatomical alterations using NDM in IGD.

Structural magnetic resonance imaging (MRI) data were obtained from 288 IGD participants and 165 matched recreational game users. Gray matter volume (GMV) was computed through CAT12 and segmented according to the Brainnetome Atlas. NMD was utilized to simulate the propagation of pathology. We initiated diffusion from each location to pinpoint probable epicenters of GMV alterations in IGD and correlated eigenmodes of the Laplacian matrix with observed atrophy and expansion patterns.

Abnormal brain regions with altered GMV were observed in IGD. Specifically, IGD demonstrated a great loss in GMV in the caudal cuneus gyrus, precentral and postcentral gyrus, as well as the cingulate cortex while simultaneously exhibiting an increase in the amygdala. The pallidus and putamen showed positive correlations with gaming craving. Both the right cingulate gyrus and the left amygdala were identified by the model as significant epicenters of disease dissemination.

The results suggest that gray matter morphological abnormalities can predict temporal sequencing of pathology progression in IGD. Subcortical gray matter volume increases in reward-processing-related regions were positively correlated with gaming craving severity in IGD, consistent with altered reward processing and motivational drive in addiction models.

Social cognitive deficits are common across many psychiatric conditions and contribute to broader social dysfunction. One hypothesized mechanism involves altered basic visual processing, which may disrupt the perception of low-level social cues and, in turn, compromise broader social cognitive processes. Here, we examined relations between basic visual processing and different levels of social cognition in a transdiagnostic youth sample.

A sample of 148 youth, ranging from healthy individuals to individuals with neuropsychiatric diagnoses and significant social dysfunction, completed two measures of basic visual processing (contrast sensitivity and visual integration) and a battery of social cognition tasks spanning lower-level (gaze perception) to mid-level (emotion recognition) to higher-level (theory of mind) social cognition. We used a four-level path model to test whether basic visual processing predicts gaze perception, which in turn predicts emotion recognition, which predicts theory of mind.

Poorer contrast sensitivity and visual integration were associated with less precise gaze perception, which was, in turn, associated with worse emotion recognition, which was associated with worse theory of mind. This four-level path model demonstrated good fit and showed superior fit compared to alternative models.

These findings suggest that basic visual processing influences the perception of basic social cues (e.g. gaze direction), which subsequently impairs more complex social perception and inference. Notably, this study extends prior observations from individuals with chronic schizophrenia to a transdiagnostic youth sample, indicating that altered basic visual processing may be a shared mechanism contributing to social cognitive deficits across psychiatric disorders and illness stages.

Smartphone-based cognitive behavioral therapy (CBT) programs offer accessible interventions for subthreshold depression, yet engagement needed for meaningful benefit remains unclear. We examined how lesson and worksheet engagement relate to depressive symptom improvements in a behavioral activation (BA) intervention, accounting for time-varying confounders.

This secondary analysis included 298 adults assigned to the BA arm of the RESiLIENT trial, a randomized controlled trial in Japan. Lesson and worksheet completion were treated as time-varying exposures, each yielding four engagement patterns: minimal (Few-Few), early (Many-Few), late (Few-Many), and consistently high (Many-Many). Outcomes were depressive symptom changes measured by the Patient Health Questionnaire-9 (PHQ-9) at weeks 6 and 26. We applied the parametric g-formula to estimate counterfactual PHQ-9 changes under each pattern, adjusting for baseline and time-varying confounders.

Early lesson engagement during weeks 0–3 was associated with larger PHQ-9 reductions at both weeks 6 and 26, even when later engagement declined (Many-Few vs. Few-Few: week 6: −1.47 [95% CI −2.52 to −0.53]; week 26: −1.27 [−2.53 to −0.17]). In contrast, higher worksheet engagement was linked to improved PHQ-9 at week 6, with maximal benefit among consistently high engagers (Many-Many vs. Few-Few: −1.25 [−2.17 to −0.44]) and late engagers (Few-Many vs. Few-Few: −1.18 [−2.20 to −0.08]), but not persist to week 26.

Greater engagement with smartphone-delivered BA is associated with larger symptom reductions. Early lesson engagement drives sustained benefit, whereas worksheet engagement did not persist. These findings may guide digital CBT design by emphasizing early lesson completion alongside concurrent skill practice.

Depression is associated with pathological dysregulations affecting both the brain and the body, with the latter being reflected in plasma proteins. While plasma protein signatures of depression have been increasingly recognized, a holistic examination of interactions with brain features is lacking.

Leveraging data from 3,966 UK Biobank participants, we identified a multimodal neuroimaging-plasma protein component of depression (NeuroPro-Dep) by integrating plasma proteins and five brain modalities via an ICD-10 diagnosis-constrained multimodal fusion approach.

Notably, NeuroPro-Dep demonstrates detectable associations with depression symptoms across datasets from diverse populations, underscoring its clinical potential. This capability is anchored in its five brain modalities alterations, including hippocampal atrophy, reduced cortical sensorimotor network functional connectivity, and impaired internetwork structural connectivity of the frontoparietal network. The multimodal neuroimaging-derived plasma protein modality of NeuroPro-Dep is enriched in metabolic pathways, as further supported by association analysis linking this modality to body mass index (BMI), type 2 diabetes, and other metabolic indicators. Crucially, two-step Mendelian randomization analysis revealed that the NeuroPro-Dep plasma protein modality exerts a causal effect on depression through BMI (plasma protein to BMI: or=0.28, p=0.035; BMI to depression: or=1.14, p=4.37×10−11).

Overall, this study underscores metabolic dysfunction as a bridge between brain changes, depression, and physical diseases, while providing a novel multimodal biological signature and valuable insights that may inform future treatment strategies.

Individuals with severe mental illness (SMI) have increased risk of physical comorbidities, linked to worse outcomes such as greater psychopathology, frailty, and neurocognitive impairment. Mechanisms underlying this burden remain unclear. This study examined whether frailty and psychopathology predict evening chronotype, especially in SMI with comorbidities.

A longitudinal study assessed 165 participants at two time points over one year: schizophrenia (n = 30), bipolar disorder (n = 42), major depressive disorder (n = 35), and healthy controls (n = 58). The SMI group (n = 107) was divided into SMI with comorbidities (SMI-C; n = 47) and without (SMI; n = 60). Measures included psychopathology, frailty, chronotype, neurocognitive and functional performance, and hematological biomarkers.

Neurocognitive and functional impairments were greater in SMI groups than controls (F = 10.3–31.4; p < 0.0001; η²p = 0.12–0.34). The SMI-C group showed worse frailty than controls at T1 (F = 4.3; p < 0.01; η²p = 0.05) and than SMI at T2 (F = 8.5; p < 0.0001; η²p = 0.12), and elevated MCV/MCH (F = 3.8–9.4; p < 0.05–0.0001; η²p = 0.04–0.11). Chronotype distribution did not differ. Frailty and psychopathology predicted chronotype in SMI (p < 0.05–0.01); in controls, frailty and performance did so (p < 0.05).

Psychopathological and hematological profiles are associated with chronotype and may help identify subgroups for chronobiology-informed interventions. These findings support more personalized treatment approaches.

Intrusive trauma memories – vivid, distressing recollections that occur involuntarily and may feel present – is a defining symptom of post-traumatic stress disorder (PTSD). While traditional models emphasize dysregulation within limbic-prefrontal circuits, emerging evidence implicates visual sensory systems in the formation of trauma-memory intrusions. However, direct causal evidence remains lacking.

We combined functional MRI with repetitive transcranial magnetic stimulation (rTMS) to examine the causal role of visual sensory cortices in intrusive memory formation. Healthy participants underwent a trauma-film paradigm, followed by fMRI scanning during memory encoding and a post-encoding resting phase during which spontaneous intrusions were recorded. One group received 1-Hz rTMS targeting early visual cortex (V1/V2); a control group received stimulation at the Vertex. Intrusive memories were recorded over the subsequent 7 days.

rTMS to V1/V2 significantly reduced the frequency, vividness, and emotional intensity of intrusive memories, while preserving recognition of episodic gist. fMRI analyses showed that intrusive episodes were associated with heightened activation and stable neural representations within the occipital visual cortex (OVC). Functional and effective connectivity analyses further revealed that the occurrence of intrusions was predicted by interactions between the middle frontal gyrus (MFG) and OVC. Dynamic causal modeling confirmed direct, bidirectional MFG-OVC interactions that coexisted with, but tracked the intrusion dynamics more closely than, the traditional prefrontal-limbic circuits.

These findings provide the first causal evidence for the direct involvement of the early visual cortex in trauma-memory intrusions. They highlight the visual system as a novel neuromodulation target for therapeutic intervention on PTSD.

Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental condition with widespread brain structure alterations. However, the relationship between macroscale cortical organization and microscale molecular mechanisms remains unclear, particularly regarding the neurobiological mechanisms shared between the full ADHD cohort and its combined subtype (ADHD-C).

We analyzed 176 patients with ADHD (105 ADHD-C, 71 ADHD inattentive subtype) and 176 matched typically developing (TD) controls from the ADHD-200 dataset. Morphometric Inverse Divergence (MIND) networks quantified cortical similarity. Partial least squares (PLS) regression linked case–control MIND differences to cortical gene expression, assessing functional enrichment, cell-type specificity, and developmental trajectories.

Neuroanatomically, the ADHD-C subtype exhibited widespread increases in regional MIND values, particularly in temporal and parietal cortices, reflecting greater inter-regional morphological homogeneity. PLS regression revealed that these MIND alterations were spatially correlated with a specific transcriptomic signature (PLS1+). These PLS1+ genes were significantly enriched in mitochondria-related metabolic pathways and showed distinct cortical layer specificity (notably layer V) and developmental stage specificity (from late fetal to late infancy stages). Regarding cell-type specificity, while PLS1+ genes in the full ADHD cohort were significantly enriched in excitatory and inhibitory neurons, the ADHD-C subtype showed similar but trend-level associations. Importantly, the full ADHD cohort and the ADHD-C group shared numerous PLS1-related genes and broad functional pathway enrichment commonalities.

This study links macroscale cortical abnormalities to microscale transcriptional regulation, with pronounced alterations in ADHD-C. The shared genetic and functional profiles between ADHD and its combined subtype underscore common pathological processes, providing novel insights into the neurodevelopmental mechanisms of ADHD.

Adverse childhood experiences (ACEs) are associated with increased risk of psychotic-like experiences (PLEs), but the relationship between specific adversities and the persistence of PLEs in young people remains unclear. We examined associations between distinct ACEs and the persistence of PLEs until 24 years old.

Using longitudinal data from participants in the Avon Longitudinal Study of Parents and Children (ALSPAC) cohort with at least one PLE datapoint, we used group-based trajectory modeling to estimate longitudinal trajectories of PLEs from age 12–24. We examined their associations with bullying victimization, maltreatment, parental mental health problems, parental substance abuse, parental separation, and parental intimate partner violence prior to first PLE experiences.

Among 4,448 participants, a three-group trajectory model provided the best fit, revealing low, increasing and persistent PLE groups from ages 12–24. In fully adjusted multinomial logistic regression models, those exposed to bullying were more likely to belong to either the increasing (relative risk ratio [RRR]: 1.83, 95%CIs: 1.26–2.66) or high (RRR: 1.78, 95%CIs: 1.07–2.93) PLEs group than the low PLE group; those exposed to maltreatment were more likely to be in the increasing PLE group (RRR: 1.47, 95%CIs: 1.03–2.10). No other ACEs were associated with PLE trajectories.

Bullying was associated with persistent PLEs up to 24 years old, independent of other forms of childhood adversity, with timing-specific effects of maltreatment on increasing symptoms emerging later in adolescence. Findings provide further evidence for the importance of prioritizing bullying and maltreatment reduction as public health targets.

Insomnia is common in major depressive disorder (MDD), and varying severity of insomnia may be associated with distinct neural alterations in MDD. Dynamic functional connectivity can capture time-varying brain network interactions and may help disentangle insomnia-related mechanisms in MDD.

We recruited 203 drug-naïve adult MDD patients, divided into high- (HI-MDD, n = 133) and low-insomnia (LI-MDD, n = 70) groups using the Hamilton Depression Rating Scale insomnia subscale, along with 122 health controls (HCs). Independent component analysis and a sliding-window approach were applied to explore static and dynamic functional network connectivity (FNC).

While static FNC revealed no significant group difference, dynamic analysis identified distinct connectivity states between two groups. Compared to HCs, HI-MDD displayed more frequent occurrence of state I and less of state IV, a pattern was absent in LI-MDD. Both MDD groups showed increased default mode network (DMN)-lateral ventral attention network (VAN) connectivity in states I and II, accompanied by decreased dorsal attention network-cerebellar/DMN connectivity in state I relative to HCs. Compared with LI-MDD, HI-MDD exhibited enhanced DMN-medial VAN connectivity in state II, along with increased DMN connectivity with visual and sensorimotor networks in state I, suggesting insomnia-related changes. In addition, we identified insomnia-related memory deficits and depression-related processing speed impairment in MDD. Insomnia significantly moderated the association between altered DMN-lateral VAN connectivity in state I and logical memory impairment in MDD.

These findings suggest that insomnia severity in MDD is associated with distinct temporal patterns of brain network alterations beyond shared depression-related changes and moderate cognitive functions in MDD.

Psychotic experiences (PEs) in are associated with elevated risk for mental health difficulties. This study examined predictors of PEs, inclusive of the role of gender, ethnicity, and protective factors.

Data were drawn from a 2021 Planet Youth survey of adolescents (n = 4,005). PEs were measured using the adolescent psychotic symptom screener. Effects of psychosocial predictors on PEs were measured by fitting multivariable logistic regression main effect and joint exposure models.

29.8% reported PEs. Black/Asian/Other minorities had elevated odds (aOR = 1.59, 95% CI 1.26–2.02, p < .001). Increased odds in males, females, undisclosed gender and non-binary/transgender with elevated emotional/behavioural difficulties (aOR = 4.47, 95% CI 3.53–5.67, p < .001; aOR = 3.25, 95% CI 2.59–4.08, p < .001; aOR = 4.83, 95% CI 2.58–9.02, p < .001; aOR = 4.33, 95% CI 2.69–6.97, p < .001 respectively). High odds in undisclosed gender with low emotional/behavioural difficulties (aOR = 4.36, 95% CI 1.50–12.66, p = .007). Lower odds from perceived school/home safety (aOR = 0.69, 95% CI 0.58–0.83, p < .001 and (aOR = 0.81, 95% CI 0.66–0.99, p = .038, respectively). Elevated odds from recent adversities (aOR = 1.91, 95% CI 1.47–2.49, p = .011) attenuated by parental support (aOR = 1.76, 95% CI 1.17–2.65, p < .001). Each additional adversity (>12 months) increased odds (aOR = 1.12, 95% CI 1.07–1.17, p < .001).

Findings highlight the interplay of risk and protective factors in adolescent PEs, with increased vulnerability among minoritized youth. Results support targeted interventions to reduce mental health disparities.

Gambling disorder (GD) involves persistent risky choices despite losses, suggesting impaired impulse control. While static paradigms reveal inhibition deficits in GD, they cannot model dynamic risk-reward escalations during real gambling. This study aims to investigate whether GD involves impaired dynamic impulse control during escalating stakes and to dissociate contributions of subjective risk evaluation and trait impulsivity to this deficit.

Using a sequential gambling task with 83 male patients with GD and 62 matched healthy controls (HCs), this study investigated dynamic impulse control deficits under escalating stakes. We quantified dynamic impulse control via the reward–reaction time (RT) coupling for ‘continue’ choices (dynamic impulse control index [DICI]) using Bayesian modeling. Risk sensitivity and risk preference were derived from stop/continue decisions. Trait impulsivity was assessed with the Barratt Impulsiveness Scale (BIS-11). Regression analyses examined the modulation of DICI by risk sensitivity and trait impulsivity.

Patients with GD exhibited significantly attenuated DICI versus HCs, reflecting failure to increase deliberation with escalating stakes. Computational modeling revealed markedly reduced risk sensitivity in GD despite comparable risk preference. Critically, trait impulsivity positively modulated DICI in HCs but not in GD, indicating pathological decoupling. Risk sensitivity positively predicted DICI in both groups, though significantly weaker in GD.

These findings establish a triadic impairment in GD: (1) attenuated adaptive impulse control during escalation (impaired DICI), (2) deficient subjective risk weighting (reduced sensitivity), and (3) breakdown of impulsivity-based modulation of control. This reveals a dynamic, mechanism-focused pathology beyond static trait models.