Background:The stress caused by multiple aspects of veterans' transitions from military to civilian, termed transition stress, represents a unique source of psychological impact that is underresearched due to its qualitative nature. The assessment of this complex psychological phenomena has thus relied on laborious interviews designed to extract quantitative information from qualitative narratives of the transition to civilian life. We sought to determine if large language models (LLMs) could be used as valid measurement tools to extract relevant information from open-ended narratives. Objective:This study sought to develop and validate a generative artificial intelligence (AI) approach to automate the quantification and subsequent thematic analysis of veteran transition stress. Methods:Utilizing transcripts from interviews of a sample of US military veterans, we developed an LLM to rate transition stress severity and examined the model's reliability in relation to human coders and validity in relation to a set of related questionnaire measures. Next, we used the LLM scores to quantitatively define high and low transition stress groups, enabling a targeted, automated analysis of themes related to narrative identity and life transition themes that might differentiate the two groups. Results:LLM ratings of transition stress correlated highly with the human expert ratings and showed significant, theoretically congruent correlations with measures of clinical symptoms, reintegration difficulties, and veterans' self-ratings of transition difficulty. Critically, the AI-derived thematic analyses of the narratives from high and low transition stress veterans revealed clearly distinct and informative patterns. Conclusions:These findings suggest that generative AI offers a robust, scalable, and reliable method for multidimensional analysis of complex, narrative-based psychological constructs.

Background: The loss of a significant other may lead to prolonged grief disorder (PGD), posttraumatic stress disorder (PTSD), depression, anxiety, somatic symptoms, and loneliness. Comparative thinking about one's well-being (e.g. social, temporal, counterfactual, or expectation-based comparisons) may act as a cognitive link between bereavement and psychopathology. For example, frequently comparing current well-being to pre-loss levels (i.e. past temporal comparison) and perceiving a gap (comparison discrepancy) can intensify negative emotions (comparison affective impact), which may, in turn, exacerbate psychopathology. We hypothesized that aversive well-being comparison frequency, comparison discrepancy, and negative comparison affective impact would each predict higher psychopathology over time, with comparison discrepancy and affective impact sequentially mediating the association between comparison frequency and psychopathology.Methods: Our two-wave longitudinal study of bereaved individuals (N = 315) assessed well-being comparisons, worry, rumination, avoidance, childhood adversity, and mental health outcomes over six months post-loss.Results: Preregistered hypotheses were partly confirmed. Higher frequency of aversive well-being comparison, comparison discrepancy, and negative comparison affective impact predicted elevated psychopathology (PGD, PTSD, depression, anxiety, somatic symptoms, & loneliness) symptoms six months later. Nuanced effects persisted after accounting for autoregressive pathways, worry, rumination, avoidance, and childhood adversity. In mediation pathways, frequent comparison amplified perceived discrepancies, eliciting negative affective impact, which in turn contributed to depression, anxiety, and somatic symptoms.Conclusion: Our work suggests that well-being comparison plays a significant role in mental health outcomes following loss.

Emotion regulation research has traditionally examined intrapersonal and interpersonal strategies in isolation, limiting understanding of their combined effectiveness in daily life. This study used ecological momentary assessment (EMA) to compare the effectiveness of combined interpersonal-intrapersonal strategies versus intrapersonal-only approaches across two samples. Study 1 analysed 10,739 observations from 156 civilians over 21 days, finding that combined strategies were modestly more effective for decreasing negative emotions but did not differ significantly from intrapersonal-only strategies for enhancing positive emotions or problem-solving facilitation. Study 2 analysed 3,894 observations from 70 veterans, revealing significantly larger combined strategy advantages for both decreasing negative emotions and enhancing positive emotions, with effect sizes approximately 2-4 times larger than civilians. Problem-solving effectiveness was unrelated to strategy type in either sample, instead being driven by the number of strategies deployed. Multilevel models revealed substantial individual variability in combined strategy benefits across both samples. These findings provide naturalistic evidence that integrating interpersonal and intrapersonal emotion regulation strategies may enhance emotional outcomes, while highlighting important population differences that have implications for personalised clinical interventions.

Research across multiple domains of investigation, from field studies to experimental laboratory work, has converged on the idea that resilience in the face of highly stressful or potentially traumatic life events is common and driven by a process of flexible self-regulation; that is, by the capacity to repeatedly adapt one’s behavior to changing situational challenges as they unfold over time. Mechanistically, flexible self-regulation is not a singular process but has been shown to involve multiple component abilities. In this paper, I will first review the literature on resilience and flexible self-regulation and how they relate to each other. I will then consider recently developed tools for the digital assessment of flexible self-regulation, specifically Ecological Momentary Assessment (EMA) and a form of artificial intelligence known as Natural Language Processing NLP). Finally, I will briefly consider potential methods to develop and enhance flexible self-regulation and, in doing so, increase the probability of resilient outcomes.

BACKGROUND:Individuals with Posttraumatic stress disorder (PTSD) can experience significant emotion dysregulation and interpersonal impairments, with emerging evidence implicating alterations in empathy as a contributing factor. Prior research links specific empathy dimensions (e.g., personal distress, perspective taking) to psychopathology; however, little is known about how patterns across empathy dimensions may map onto distinct profiles, which in turn may be related to different clinical outcomes. The identification of empathy profiles in PTSD could provide the opportunity to tailor trauma interventions to specific patterns of empathic responding and improve interpersonal functioning. Therefore, this study aimed to identify empathy profiles in PTSD and their associations with emotion dysregulation and related clinical symptoms. METHODS:Participants were 108 adults with PTSD (77.8% female; mean age = 38.76) seeking trauma-focused treatment. K-means cluster analysis on the Interpersonal Reactivity Index empathy subscales identified empathy profiles. Linear models compared profiles on clinical measures (childhood trauma, depression, anxiety, PTSD symptoms, dissociation, emotion dysregulation). RESULTS:Three profiles emerged: High Empathy/Moderate Distress (32.4%), Moderate Empathy/Low Distress (38.0%), and Low Perspective-Taking/High Distress (29.6%). Follow-up univariate tests showed profiles differed significantly in depression, anxiety, and emotion dysregulation but not trauma history, PTSD symptoms, or dissociation. Emotion dysregulation differences remained significant after controlling for clinical and demographic covariates. CONCLUSIONS:We identified novel empathy profiles in PTSD that show differential associations with emotion dysregulation and psychopathology. Emotion dysregulation emerged as the primary clinical correlate distinguishing these profiles, suggesting it may represent a key mechanism linking empathic patterns to functional outcomes.

This prospective study examined whether exposure to potentially morally injurious events (PMIEs), specifically self-attributed transgressions, other-attributed transgressions, and experiences of betrayal, predicted change in posttraumatic stress symptoms (PTSS) and reintegration difficulty during the military-to-civilian transition (MCT). U.S. Army soldiers (N = 815) completed surveys approximately 6 months before separation (Time 1 [T1]) and 6 months after separation (Time 2 [T2]). PMIEs were assessed at T1, PTSS was assessed at both points, and reintegration difficulty was assessed at T2. Analyses used hierarchical linear regression and moderation models, adjusting for demographic and service-related covariates. Betrayal-related PMIEs uniquely predicted higher PTSS at T2, β = .09, p = .014, after controlling for baseline PTSS, whereas self- and other-attributed transgressions were nonsignificant. Betrayal-related PMIEs also predicted higher levels of reintegration difficulty above and beyond the contributions of PTSS and combat exposure, β = .13, p = .003. Interaction effects by gender and relationship status were statistically significant but small, ΔR2 = .015. Across models, effect sizes for betrayal-related PMIEs were modest, βs = .09-.15. Because the PTSD Checklist-Military Version was administered without anchoring responses to a specific traumatic event, the findings reflect general stress-related distress rather than DSM-defined PTSD. Overall, the results indicate that betrayal-based PMIEs represent a modest yet consistent risk factor for psychological and functional difficulties during the MCT, underscoring the value of differentiating PMIE subtypes in screening and intervention.

Predicting treatment non-response for anxiety and depression is challenging, in part because of sparse symptom assessments in real-world care. We examined whether passively captured, fine-grained emotions serve as linguistic markers of treatment outcomes by analyzing 12 weeks of de-identified teletherapy transcripts from 12,043 U.S. patients with moderate-to-severe anxiety and depression symptoms. A transformer-based small language model extracted patients' emotions at the talk-turn level; a state-space model (VISTA-SSM) clustered subgroups based on emotion dynamics over time and produced temporal networks. Two groups emerged: an improving group (n=8,230) and a non-response group (n=3,813) showing increased odds of symptom deterioration, and lower likelihood of clinically significant improvement. Temporal networks indicated that sadness and fear exerted most influence on emotion dynamics in non-responders, whereas improving patients showed balanced joy, sadness, and neutral expressions. Findings suggest that linguistic markers of emotional inflexibility can serve as scalable, interpretable, and theoretically grounded indicators for treatment risk stratification.

Recent research has challenged the assumption that specific emotion regulation (ER) strategies are universally adaptive, emphasizing instead ER flexibility-the ability to adjust strategies in response to shifting situational demands. Contemporary theory conceptualizes ER flexibility as a multicomponent process involving context sensitivity, access to an effective strategy repertoire, and feedback responsiveness. Although ER flexibility is positioned as key to adaptation following traumatic events, empirical work has largely focused on normative, nonclinical samples, leaving real-world affective benefits in trauma-exposed populations unclear. Using ecological momentary assessment, we examined momentary associations between ER flexibility components and momentary psychopathology symptoms in 64 trauma-exposed U.S. veterans (4,371 observations), including 27 who met posttraumatic stress disorder (PTSD) diagnostic criteria. Participants completed four surveys per day for 21 days assessing emotional situations, situational characteristics, ER strategy use and change, and momentary symptoms. Multilevel models tested within-person associations between flexibility components and symptoms and examined whether effects differed by PTSD status. Higher context sensitivity, β = -.09; repertoire use, β = -.05; and feedback responsiveness (strategy switching and initial effectiveness interaction), β = .10, were independently associated with lower momentary psychopathology symptoms, ps < .05. Within feedback responsiveness, ps < .01, maintaining initially effective strategies was associated with lower anxiety, β = .27, whereas switching from ineffective strategies was associated with lower depression, β = -.17. Associations did not differ by PTSD status. ER flexibility supports adaptive emotional functioning in daily life among trauma-exposed veterans regardless of PTSD diagnosis, highlighting its potential clinical relevance as a transdiagnostic process.

The increasing reference by young people on social media to thoughts of suicide or self-harm through "trauma-dumping" posts has prompted platforms like TikTok to ban the use of suicide-related language. Youth are often able to evade this ban using veiled "algospeak" phrases, such as "unalive," "sewerslidal," and "SH." This study assessed the potentially harmful or supportive nature of such content. We hypothesized that videos using algospeak would feature greater harmful content compared to videos without algospeak. Transcriptions of 60 videos were generated using Whisper AI. Language and the presumed purpose of the video were coded by both humans and artificial intelligence (AI) using a valence Likert scale from 1 (harmful) to 5 (supportive). Human and AI coders were highly correlated (algospeak: r = .95, p < .01; nonalgospeak: r = .91, p < .01). For both human and AI coders, mean nonalgospeak scores were significantly higher (p < .05) and in the supportive direction (M-human = 3.83, SD = 0.96; M-AI = 3.60, SD = 1.19) than algospeak content (M-human = 2.95, SD = 1.05; M-AI = 3.00, SD = 1.05). These results serve as preliminary evidence suggesting that videos using the algospeak feature have more harmful suicide or self-harm-related content compared to videos using conventional language (nonalgospeak).

Refugees are often forced to leave their homes and rebuild their lives in unfamiliar environments. Adapting to a new place can be challenging, as it involves coping with the loss of familiar surroundings while navigating new ones. Emotional bonds with places-known as place attachment-may play an important role in this adaptation process, serving both as a protective factor and a potential burden. We examined these dynamics in a study among Ukrainian refugees in Poland (N = 1,016) following the full-scale Russian invasion. We measured attachment to participants' hometowns in Ukraine and their current towns of residence in Poland, as well as life satisfaction, symptoms of post-traumatic stress disorder (PTSD), and post-traumatic growth (PTG). Stronger attachment to one's hometown in Ukraine was associated with higher PTSD symptoms and lower life satisfaction. In contrast, attachment to a new town of residence in Poland was positively linked to life satisfaction and PTG. Attachment to the former hometown was negatively related to attachment to a new place in Poland. However, the more similar the two locations were perceived to be, the stronger the attachment to the new town. These findings suggest that place attachment can act as both a resource and a risk factor, depending on how migrants navigate continuity between past and present places.

Objective: The current study compared symptom networks between individuals exhibiting resilience and nonresilience trajectories of adaptation two years after the COVID-19 outbreak. Method: A population-representative sample (N = 906) reported symptoms of anxiety and depression in February-July 2020 (T1), March-August 2021 (T2), and September 2021-February 2022 (T3), as well as symptoms of post-traumatic stress disorder (PTSD) and adjustment disorder (AD) at T3. After differentiating between individuals with resilience and non-resilience trajectories using growth mixture modeling, network analyses were conducted to investigate group differences in T3 network symptoms (undirected and directed). Results: Despite non-significant group differences (M = 0.184, p = .380; S = 0.096, p = .681), distinctive qualitative characteristics were observed between networks. Difficulty relaxing was identified as the single root cause in the more diffused resilience network, with anxiety and depressive symptoms as additional starting points in the non-resilience network, which was more interconnected into clusters with clear-cut diagnostic boundaries. Sad mood demonstrated a transdiagnostic communicative role across common mental disorders. Conclusion: Our results contribute to the understanding of anxiety-depression-PTSD-AD symptom networks in resilient and non-resilient individuals by highlighting the consequences of heterogeneity in adaptation capacity in the development of pandemic-related psychopathology.

CONTEXT:As one of the largest-scale public health disasters, the COVID-19 pandemic has significantly disrupted daily routines and resulted in poorer mental health especially among more socioeconomically disadvantaged populations. OBJECTIVE:In this study, we investigated trajectories of routine disruptions and their relationships with mental health symptom trajectories, as well as socioeconomic characteristics between 2020 and 2022 amid the COVID-19 pandemic. METHODS:A population-representative sample in Hong Kong (N = 1333) was recruited to complete self-report instruments at the pandemic's acute phase (February-July 2020, T1), and again at 1-year (March-August 2021, T2) and 1.5-year (September 2021-February 2022, T3) follow-ups. Respondents reported primary and secondary routine disruptions, and depressive and anxiety symptoms. RESULTS:Growth mixture modeling (GMM) identified four trajectories of routine disruptions: sustained regularity (55.66 %-74.57 %), recovery from disruptions (5.10 %-14.40 %), delayed disruptions (10.20 %-14.03 %), and chronic disruptions (6.30 %-19.73 %). Four symptom trajectories were also demonstrated: resilience, recovery, delayed distress, and chronic distress. Respondents showing sustained regularity and recovery from disruptions were more likely to be dependent (vs. employed) and demonstrate a resilience or recovery trajectory on mental health symptoms. Respondents showing chronic and delayed disruptions were more likely to be younger with lower socioeconomic status and demonstrate a chronic or delayed distress trajectory on mental health symptoms. CONCLUSION:People's routine disruptions displayed heterogeneous trajectories over time. Sustained regularity was the most prevalent among others. Access to socioeconomic resources could dictate optimal behavioral adjustment. Patterns of daily routine disruptions demonstrate a concordance with patterns of psychological adjustment. These results provide an initial evidence base for advancing the nature and role of behavioral adjustment during and after large-scale disasters.

The current study examines trajectories of distress symptoms across five time points during the 10-month "Iron Swords" War between Israel and Gaza. Participants (N = 957) were adult Israeli Hebrew speakers who responded to all five assessments. Latent growth mixture modeling indicated a three-class trajectory model: (a) resilience (59%) characterized by mild initial distress that decreased slightly but steadily over time; (b) moderate-stable (33%) characterized by moderate initial distress that remained relatively unchanged; and (c) emerging-chronic (8%) characterized by relatively high initial distress that steadily increased over time. Logistic regression in a conditional model indicated that participants in the resilient group were less likely to be female, single/divorced, and more likely to be orthodox and supportive of the government than those in the moderate-stable group. Resilient participants were also less likely to be female and more likely to be married than those in the emerging chronic group. Finally, the participants in the emerging-chronic group were less likely to be married than those in the moderate-stable group.

Depression, a common outcome following stroke, has been repeatedly associated with poststroke cognitive deficits and is inversely associated with prestroke cognitive functioning or cognitive reserve. In particular, affected executive functioning has been linked to poststroke depression severity whereas preserved executive functioning is associated with poststroke recovery. Moreover, a growing body of research has demonstrated distinct prototypical trajectories of depression following major medical events. This work has consistently indicated that resilience, or a stable trajectory of healthy adjustment following adverse or potentially traumatic events, is the most common outcome. To examine trajectories of poststroke depression and their associations with prestroke cognitive functioning. Secondary longitudinal data analysis of a large cohort study. Institutional; RAND Health and Retirement Study sponsored by the University of Michigan and the National Institute on Aging. Participants (N = 2298) were drawn from a national, longitudinal dataset of older adults who experienced a stroke and survived for at least 2 years afterward. Not applicable. Depression was assessed via a modified version of the Center for Epidemiologic Studies Depression Scale Short Form. Results showed that resilience was the most common outcome (65.5% of the sample) following stroke and was associated with greater prestroke working memory (p < .001), age (p < .001), and gender (p < .001). Stroke survivors demonstrate the same depression trajectory patterns as observed following other major medical events. These results help illuminate the longitudinal relationship between working memory and poststroke depression and potentially inform interventions. Additional research is required to understand how these findings may be applied in clinical settings.

Previous research indicates that many veterans experience challenges when transitioning from military to civilian life, with a significant proportion reporting feelings of social exclusion and loneliness. However, these experiences have been documented primarily through self-report surveys with limited examination of their inter-relationship. The current study addressed this gap using an experimental Cyberball paradigm designed to simulate social exclusion. Specifically, we investigated the relationship between perceived exclusion and loneliness among post-9/11 U.S. military veterans. Veteran participants (N = 191) were randomly assigned to one of four conditions in a 2 (inclusion vs. exclusion) x 2 (veteran vs. nonveteran online confederates) between-subjects design. Results revealed a significant interaction effect, with veterans reporting higher levels of loneliness when excluded by nonveteran confederates compared to when excluded by veteran confederates. Nostalgia for the military and warrior identity were each significantly associated with loneliness but did not moderate the effect of exclusion. These findings provide experimental evidence for the impact of perceived nonveteran exclusion on veteran loneliness and highlight the importance of facilitating veterans' social integration during the transition to civilian life.

Emotion regulation (ER) plays a central role in mental health, but the effect differs across cultures. Here, expanding from extant literature's focus on Western-Eastern dichotomy or individualism-collectivism, this meta-analysis synthesized evidence on the associations between the two most-studied ER strategies (cognitive reappraisal and expressive suppression) and two mental health outcomes (psychopathology and positive functioning) and investigated the moderating roles of several cultural dimensions: Hofstede's national cultures dimensions, education, industrialization, richness and democracy (EIRDness), and sample demographics. A comprehensive literature search was conducted using electronic databases (CINAHL, Scopus, Web of Science, PsycINFO and MEDLINE) to identify eligible studies reporting relationships between ER and mental health outcomes (PROSPERO: CRD42021258190, 249 articles, n = 150,474, 861 effect sizes, 37 countries/regions). For Hofstede's national cultures and EIRDness, multimodel inference revealed that greater reappraisal propensity was more adaptive in more short-term-oriented, uncertainty-tolerant and competition-driven cultures, whereas greater suppression propensity was more maladaptive in more indulgent and competition-driven cultures. For demographics, greater reappraisal propensity was more adaptive for samples with more female (B = -0.19, 95% confidence interval (CI) -0.29 to -0.09) and more racial minority participants (B = -0.32, 95% CI -0.51 to -0.13), whereas greater suppression propensity was more maladaptive in younger samples (B = -0.004, 95% CI -0.005 to -0.002). These findings elucidate how cultures are associated with the function of ER and suggests ways in which future studies can integrate cultural characteristics when examining ER and psychological adjustment.

The first two years of COVID-19 triggered a significant and widespread set of stressors, which disproportionately affected minoritized populations. This study utilized the Boston College COVID-19 Sleep and Well-Being Dataset to examine changes in depression, anxiety, and sleep disturbance during a collective stressful event (CSE). Data was examined over four periods spanning from May 2020 to October 2021. Latent growth mixture modeling (LGMM) was used to identify discrete growth trajectories of the outcome measures and test for covariates associated with each trajectory. Two distinct symptom trajectory classes emerged for both sleep quality and anxiety (chronic symptoms and resilience) and three classes for depression (recovery, delayed symptoms, and resilience). Neuroticism was inversely associated with resilience to depression, anxiety, and sleep difficulties. Sexual minority (SM) individuals were more likely to be categorized in the chronic trajectory for sleep difficulties and depression. These findings suggest that during periods of widespread chronic stress, SMs might experience disproportionate negative mental and behavioral health outcomes.

The assessment of complex psychological phenomena has traditionally relied on resource-intensive qualitative methods or self-report measures that may not capture nuanced experiences. This study introduces and validates Generative Artificial Intelligence (GenAI) approach to automate the quantification and subsequent thematic analysis of veteran transition stress. Utilizing transcripts from interviews, a Large Language Model rated transition stress severity. These ratings correlated very highly with transition stress ratings from human experts and showed significant, theoretically congruent correlations with measures of clinical symptoms, reintegration difficulties, and veterans' self-ratings of transition difficulty. Critically, the AI-derived scores were used to quantitatively define high- and low-stress groups, enabling a targeted, automated thematic analysis. This analysis revealed distinct narrative structures of resilience and chronic stress characterized by themes of deployment, loss/trauma, material/emotional support and strategies. These findings suggest that GenAI offers a robust, scalable, and valid method for multidimensional analysis of complex, narrative-based psychological constructs.

Despite extensive research examining the associations between sleep and depression, there is limited evidence on how emotion regulation (ER) may mitigate their day-to-day relationships. This study aimed to investigate whether ER repertoire, defined as the breadth of strategies a person can access, moderates the relationship between sleep quality and next-day depression during a collective stressful life event, the COVID-19 pandemic. Data were drawn from 389 participants (14,457 observations) from the Boston College Daily Sleep and Well-Being Survey study. Participants completed baseline assessments of ER strategies (positive reappraisal, refocusing on planning, acceptance, putting into perspective, positive refocusing) and at least 20 daily diary entries reporting depression and prior-night sleep quality (awake time, sleep efficiency, and subjective difficulty with sleep). A novel computational formula was developed to quantify ER repertoire across multiple strategies. Multilevel modeling was used to examine within-person and between-person associations. Analyses indicated that next-day depression was associated with both within-person and between-person variations in sleep quality. ER repertoire moderated the between-person, but not within-person, associations between depression and two sleep variables (awake time and sleep efficiency). Individuals with a larger ER repertoire experienced greater emotional benefits from efficient sleep and were less adversely impacted by extended awake time. Findings highlight the importance of targeting both sleep quality and ER repertoire in interventions designed to promote emotional well-being during adverse life events. Enhancing ER repertoire may strengthen resilience against the emotional consequences of poor sleep.

Recent research in coping and emotion regulation emphasizes the importance of flexibly deploying strategies to meet situational demands. Because flexible responding is complex and performed under stress, motivation to selfregulate is theorized to be important to its success. A flexibility mindset has been proposed as a source of this motivation. The current study investigates the theorized motivational role of a flexibility mindset by testing whether components of the mindset moderate the relationship between flexibility skills and mental health symptoms. As part of an observational, cross-sectional study in 2019, Adult MTurk workers (N = 802) completed measures of flexibility skills, flexibility mindset components, and psychopathology. We tested whether two proposed components of a flexibility mindset, optimism and self-efficacy, moderated between previously established latent profiles of flexibility skills and both depression and anxiety. Supporting the hypothesized role of the flexibility mindset, both optimism and self-efficacy moderated between flexibility and depression, though optimism was more robust. Flexible regulators showed larger decreases in symptoms as optimism increased. Optimism, but not self-efficacy, also moderated the relationship between flexibility and anxiety. In this case, though, unskilled regulators showed larger decreases in symptoms as optimism increased. These findings offer an initial test of the flexibility mindset theory.