Search Results (23,787)

Influence maximisation traditionally assumes that each activated neighbour contributes independently to the likelihood of a user adopting information, ignoring conjunctive synergies where a set of users must be active simultaneously to trigger another. Relational games provide a formal language for such coalitional dependencies, but the necessary influence hypergraph must be hand-crafted by domain experts, making them infeasible for large, dynamic social networks. We introduce the Neural Probabilistic Relational Game (N-PRG), a data-driven framework that automatically learns a probabilistic hypergraph of influence coalitions from cascade traces. A feed-forward neural network, trained via gradient descent to predict user activation, is interpreted using Deep SHAP to extract important set-level triggers. These are calibrated into a stochastic cascade model, the Probabilistic Relational Game (PRG), which generalises the Independent Cascade to set-based activation. We define the Minimal Reliable Seed Set problem, prove its NP-hardness even in the deterministic case, and establish that the expected influence function is monotone. We further demonstrate that, unlike the Independent Cascade model, the influence function under conjunctive (AND-type) hyperedges is in general not submodular, which precludes constant-factor approximation guarantees and motivates the use of greedy heuristics. Extensive experiments on synthetic data confirm that N-PRG successfully identifies coalitional interactions of size greater than one and achieves targeted out-of-sample coverage. Semi-synthetic experiments on Digg and Twitter network topologies demonstrate that N-PRG discovers seed sets up to 45% smaller than Independent Cascade baselines, while providing interpretable coalition pathways invisible to black-box methods. N-PRG thus unites the flexibility of gradient-descent learning with the structural rigour of relational games for influence analysis. Full article

Interaction-level activation patching is used to infer nonlinear cooperation in neural networks, but its numerical validity has not been systematically audited. In two pretrained checkpoints—GPT-2-medium and Pythia-410M, audited on one CUDA backend with a single templated cloze task family—we identified two independent failure modes. First, exact finite pair interactions subtract four $O\left(1\right)$ forward evaluations while the target scales as $O\left({\alpha }^2\right)$, creating a precision floor. In matched float32/float64 experiments, at a descriptive $3\times$ floor, 13.0% and 51.2% of full-displacement interactions fell below the threshold—across reasonable $2\times$–$10\times$ cutoffs, these fractions ranged from 8.6 to 33.0% and from 39.5 to 82.1%, respectively—with sign disagreements of 2.9% and 13.2%; deterministic repetitions showed zero spread and therefore failed to expose the error. Second, even in float64, endpoint cross-Hessian estimates diverged from finite interactions as displacement increased, reaching 41–68% error at the full-replacement scale commonly used in patching; this displacement relationship is fitted on only these two checkpoints, and the Pythia-410M fit is visibly less stable. The corrupted prompts contain neither candidate answer, and the fixture’s behavioral contrast was not serialized at measurement time. A post hoc audit of the exact deposited fixture now confirms the intended contrast (the answer beats the distractor on all clean prompts in both models; the clean-minus-corrupt contrast is positive on $20/20$ and $17/20$ prompts). Thus, the quoted constants are properties of this behaviorally supported fixture and backend; we expect the audit procedure, not the constants, to transfer. We introduce an inexpensive ${\alpha }^2$ scaling audit that detects cancellation floors and hidden mixed-precision bottlenecks; it exposed a hardcoded float32 softmax path inside nominal-float64 GPT-NeoX inference. A motivating negative result—that local logical gate topology does not predict attribution-patching error—was obtained under an author-held internal specification that is not independently timestamped and only in ten small synthetic transformers; it is untested at pretrained scale. These results show that reproducibility alone does not establish measurement validity and that precision error and endpoint approximation error must be audited separately. The proposed checks provide a practical validation standard for interaction-level mechanistic interpretability claims. Full article

Introduction: Mobile phone text messaging (Short Message Service; SMS) interventions are digital health strategies that may support mental health care. This scoping review examined related study designs, outcomes, and gaps across the literature. Methods: Systematic searches identified 38 eligible studies from 1542 abstracts. Data were extracted and thematically analyzed to identify trends and outcomes. Texting interventions were categorized as cognitive, motivational, instrumental/practical, or combined approaches. Results: Study design included randomized controlled trials (RCTs), quasi-experimental pilots and systematic reviews, with RCTs providing the strongest evidence for improvements in adherence and relapse prevention. Outcomes included behavioral measures (e.g., medication adherence, therapy homework completion, and reduced non-attendance) alongside clinical measures (e.g., symptom reduction and relapse prevention). Cognitive reminders that were most common reduced non-attendance and improved adherence, particularly for medication and appointments, although they were less effective in severe depression. Motivational and combined interventions demonstrated higher acceptability and engagement, particularly in adolescents and high-risk individuals. Success was influenced by diagnosis, age, gender, and social status, including adherence at baseline. Conclusions: Overall, text-based interventions were generally acceptable, accessible, and cost-effective for improving mental health engagement. Limitations included lack of long-term evaluations, technological barriers, and lack of personalization. Future research should address these gaps and include underrepresented populations. Full article

Carbon-fiber structural batteries represent a class of multifunctional energy-storage systems that integrate electrochemical energy storage with mechanical load-bearing capability. Unlike conventional batteries, which are mainly evaluated based on cell-level energy density, structural batteries provide new opportunities for system-level weight reduction by reducing inactive structural mass, improving space utilization, and enabling distributed energy storage within integrated structures. In recent years, substantial progress has been achieved in carbon-fiber electrodes, structural electrolytes, laminated devices, electrolyte topology engineering, and fully carbon-fiber structural batteries. Nevertheless, most reported advances have been demonstrated under relatively ideal static testing conditions, while maintaining multifunctional performance under manufacturing and long-term service conditions remains a critical challenge. This review systematically examines the development of carbon-fiber structural batteries from a reliability perspective. First, the system-level motivations and technological evolution are introduced, and existing architectures are categorized according to their integration depth and degree of multifunctional coupling. Carbon-fiber electrodes are then discussed with emphasis on balancing capacity, ion transport, cycling stability, mechanical property retention, interfacial robustness, and manufacturing scalability. Furthermore, structural electrolytes are reviewed from the viewpoint of topology-enabled regulation of ion transport and load transfer, with particular focus on the intrinsic trade-off between ionic conductivity and mechanical modulus. In addition, manufacturing routes and device architectures are analyzed from the perspective of multifunctionality-degrading defects, including voids, dry regions, coating cracks, weak interfaces, and current-collector discontinuities. Finally, retained multifunctionality is used as a reliability-oriented evaluation criterion to examine the preservation of electrochemical, mechanical, interfacial, and safety functions, with particular emphasis on the carbon-fiber-specific failure chain linking interfacial and manufacturing heterogeneities to multifunctionality-degrading defects, coupled-field localization, and damage propagation. This review emphasizes that reliable carbon-fiber structural batteries require application-specific and coordinated optimization of materials, interfaces, electrolyte topology, coupled degradation behavior, and validation protocols. Full article

Background: Creatine is now discussed beyond sports performance and within women’s health, cognition, fatigue, and healthy-aging contexts. However, little is known about the characteristics of women who use creatine or whether these characteristics differ across durations of use. This study characterized women reporting current or past-12-month creatine use in a large cross-sectional sample and benchmarked findings against men. Methods: In a cross-sectional, supplement-platform convenience survey (approximately 2.0% response rate; a secondary analysis of a previously reported cohort), 2094 current or recent creatine users (1055 women, 1039 men; mean age 53 years [54.0 in women and 52.5 in men]) reported duration of use, supplementation goals, information sources, creatine knowledge alignment using a five-item battery, self-rated confidence, satisfaction, and intent to continue supplementation for another six months. Representation of women across six ordered duration bands was evaluated using an age-adjusted logistic model with duration as a categorical factor, with a Cochran–Armitage trend test reported alongside it. Among women, goals and information sources were compared across duration-of-use bands using unadjusted tests of homogeneity, and knowledge alignment was evaluated using an ordered trend test. Gender differences were estimated in age- and duration-adjusted models, pooled across duration groups, with Benjamini–Hochberg correction within domain; gender × duration interactions were tested for the goal and information-source outcomes. Self-rated confidence was modeled as a function of knowledge, gender, and their interaction. Results: Women comprised 50.0–63.9% of respondents across each reported duration category below three years but only 19.7% of those reporting 3+ years of use. In the primary age-adjusted analysis, gender composition differed strongly across duration categories (global p < 0.001); relative to the 3+ year group, the odds of identifying as a woman were 3.9- to 7.5-fold higher in every shorter-duration category. Among women, knowledge alignment was higher at longer reported durations in the unadjusted analysis and remained associated with duration after adjustment for age, BMI, and physical activity (global categorical test, p < 0.001). Lower knowledge alignment among shorter-duration women reflected greater uncertainty rather than greater myth endorsement, which remained rare across duration groups. Compared with men, women were less likely to report athletic performance as a goal (adjusted 37.5% vs. 50.2%) and more likely to report healthy aging, while six of seven information sources differed by gender. Women also had lower knowledge-alignment scores overall, yet the largest gender difference emerged in confidence: at equivalent observed knowledge alignment, women reported progressively lower self-rated confidence than men as scores increased (interaction p = 0.012), with an adjusted difference of −0.38 points at the highest observed knowledge score. Conclusions: Among current and recent SuppCo users who responded to this survey, women showed a pattern of creatine use that differed from the historically performance-focused populations underlying much of the evidence base. Women were disproportionately represented in shorter reported duration-of-use categories, more often endorsed healthy aging and less often athletic performance, and reported different information sources than men. Lower knowledge alignment among shorter-duration women reflected greater uncertainty rather than greater endorsement of misconceptions, while women reported lower confidence than men at the highest observed knowledge levels. These findings suggest future research and education should go beyond efficacy trials in women to address who is using creatine, the outcomes they expect, where they obtain information, and how knowledge and confidence shape decision-making. Prospective studies are needed to determine what underlies the difference in women’s representation across reported duration-of-use groups. Full article

Home-based physiotherapy is performed without supervision, which leads to incorrect execution and motivates systems that assess movement automatically from inertial measurement units (IMUs). Such systems assign each repetition to a category, yet a relevant share of repetitions fall near a class boundary, where even trained raters disagree. Classifiers trained with one-hot labels collapse these borderline repetitions onto a single class and discard this ambiguity. To address this, we build on label distribution learning, which represents each repetition as a distribution over classes instead of a single label. We introduce a way to construct such distributions without a large rater pool by perturbing the thresholds of a rule-based evaluation procedure to simulate rater disagreement. We train a network to reproduce these distributions with a Kullback–Leibler objective, which we call the ambiguity approach, and compare it against a one-hot cross-entropy baseline on four IMU exercise datasets. From the predicted distribution we then determine whether a repetition is ambiguous and which classes are relevant to it. The ambiguity approach matched or exceeded the baseline classification on all four datasets and detected ambiguity and the relevant classes more reliably. Representing the label distribution in the training target therefore adds information about ambiguity at no cost to classification. Full article

Supraventricular tachycardias (SVTs), including atrial fibrillation (AF), atrioventricular nodal re-entrant tachycardia (AVNRT), atrioventricular re-entrant tachycardia (AVRT), and focal atrial tachycardias, can possibly arise from the interaction of genetic predisposition and environmental exposures. While genome-wide association studies (GWASs) have identified 525 loci for atrial fibrillation and a small number of loci for AVNRT and accessory pathway-mediated tachycardia, the contribution of air pollution, lifestyle factors and psychosocial stress to epigenetic remodeling of atrial tissue remains insufficiently integrated into current mechanistic models of arrhythmogenesis. This review aims to synthesize evidence on how environmental exposures, including PM_2.5, NO₂, ozone, tobacco smoke, obesity, alcohol use, and physical inactivity, modulate epigenetic pathways relevant to SVT susceptibility, and to evaluate whether AF-derived epigenetic insights can be extrapolated to other SVTs. Thus, a narrative synthesis was conducted across studies examining environmental determinants, epigenetic mechanisms (DNA methylation, histone modifications, non-coding RNAs), and genetic susceptibility in SVTs. Literature from cardiac tissue studies, circulating epigenetic biomarker analyses, and mechanistic AF models was integrated to construct a unified gene–environment–epigenome framework. In atrial fibrillation, environmental exposures are consistently associated with epigenetic alterations affecting atrial electrophysiology, inflammation, oxidative stress and structural remodeling, and air pollutants and lifestyle factors modulate methylation signatures, histone-modifying enzymes and microRNA networks implicated in atrial conduction and re-entry. For AVNRT, AVRT and focal atrial tachycardia the evidential position is different. Large prospective cohort and case-crossover analyses now link air pollution to incident and acute supraventricular tachycardia, and genome-wide association studies have identified susceptibility loci for AVNRT and for accessory-pathway-mediated tachycardia, including one gene encoding a cardiac chromatin-remodeling protein; but no epigenomic profiling of nodal or accessory-pathway tissue has been reported, and no study has measured an environmental exposure, an atrial epigenetic mark and a non-AF SVT endpoint in the same participants. Twin data indicate that approximately 35% of SVT risk is attributable to genetic and 65% to unique environmental factors, which motivates a gene–environment–epigenome framework without validating its mechanistic detail outside AF. Mapping GWAS-identified loci onto environmentally responsive regulatory pathways identifies candidate convergence points between inherited risk and exposure-driven remodeling; for non-AF SVT, these are designated working hypotheses rather than established mechanisms. Air pollution and lifestyle factors are associated with supraventricular arrhythmia across the spectrum, and in atrial fibrillation there is direct evidence that they act, in part, through epigenetic reprogramming of atrial tissue. No epigenetic panel has been prospectively validated for the prediction of any supraventricular arrhythmia, and precision risk stratification therefore remains a research objective rather than a near-term clinical horizon. Integrating environmental exposure data with genetic and epigenomic profiling is nonetheless the most plausible route toward it. Future priorities include exposure-stratified, cell-resolved epigenomic profiling of atrial and nodal tissue, prospective validation of candidate circulating markers against incident arrhythmia, and replication in non-European populations and in both sexes. Full article

Background/Objectives: There is limited research on clinical profiles of school-aged children receiving a first-time diagnosis of autism spectrum disorder (ASD) Level 1 and mixed findings related to sex differences in phenotype. This study aimed to describe the clinical profile of school-aged children receiving a first-time diagnosis of ASD Level 1, explore sex differences in phenotype by symptom severity and symptom domain profiles, and compare parent and teacher ratings on standardized measures of social, emotional, and behavioral functioning. Methods: A retrospective chart review of an ASD assessment clinic was completed. Eighty-one school-aged children were diagnosed with ASD Level 1. Measures of social, emotional, adaptive, and behavioral functioning were completed. Results: Females were significantly older at the time of diagnosis (M = 9.7 years) in comparison to males (M = 8.5 years). Additionally, over half of children diagnosed with ASD Level 1 presented with ADHD and over a third presented with an anxiety disorder. Approximately one-fourth of those diagnosed were currently taking psychotropic medication, and a substantial proportion had reported speech or language delay. Approximately half presented with food selectivity and sleep problems, with only 38% currently receiving behavioral health therapy services. Caregivers reported significantly higher internalizing symptoms and externalizing behavior on all Achenbach scales in comparison to teachers. Caregivers also reported significantly greater autism-specific social concerns. No statistically significant sex differences were found in parent or teacher ratings of domain scores. Based on caregiver reports, males demonstrated greater severity of aggressive behavior. Females displayed greater deficits in social communication and motivation based on both parent and teacher reports. Overall cognitive ability fell in the average range, while adaptive behavioral functioning was in the moderately low range. Conclusions: ASD Level 1 in school-aged children presents a distinct clinical profile marked by high rates of co-occurring conditions (ADHD and anxiety), existing academic accommodations/modifications, physician referral for evaluation, and varied symptom severity by informant type, all of which may contribute to diagnostic overshadowing and delay, especially in females. Full article

Automated surface defect detection in hot-rolled steel is a prerequisite for real-time quality control, yet most deployed inspection systems operate in isolation from Industrial Internet of Things (IIoT) infrastructure. This paper reports a laboratory-scale proof of concept with two contributions. The first is a segmentation study on the Severstal dataset using a leakage-free, defect-stratified split of 1886 test images. Because a trivial all-background predictor already attains 96.66% pixel accuracy, performance is reported through Dice, IoU, precision, recall, and F1 with 95% confidence intervals. A compact from-scratch U-Net (0.49 M parameters) reaches a Dice of 0.416 at 38.6 ms per image, an ImageNet-pretrained DeepLabV3+ model reaches 0.677 at 46.3 ms and 37 times the parameters, and a classical Otsu baseline reaches 0.060, bracketing an explicit accuracy-versus-footprint design space rather than a single recommended model. The second contribution is architectural: a three-layer IIoT architecture whose messaging layer is empirically characterized on a Raspberry Pi broker over 158,500 messages. A factorial experiment isolates the transport configuration of the broker, rather than that of the publisher, as the determinant of end-to-end latency, yielding a seventeen-fold reduction. The layer sustains 1920 messages per second without loss, and a deliberate broker outage shows that MQTT delivery guarantees are semantic rather than temporal, motivating an application-level message-expiry policy. Embedded inference deployment is identified as the primary next step. Full article

Legionnaires’ disease incidence has risen substantially across industrialised nations since the early 2000s, yet standard control protocols—primarily secondary disinfection verified by culture—rely on an evidence base that warrants further scrutiny. This review offers three interpretations of the available evidence, without claiming they are established conclusions. First, culture-based verification has inherent methodological limitations: treatments that reduce culturability can induce viable-but-non-culturable (VBNC) states undetected by culture alone, creating potential discrepancies between verified culture reduction and complete microbial clearance. This reading rests on evidence that monochloramine at ≥1 mg/L abolishes culturability while genomic units remain essentially unchanged and that starved VBNC Legionella retain the capacity to infect human macrophages after more than 220 days, albeit at roughly 1% the efficiency of culturable cells; whether this translates into material risk in operating buildings remains undetermined. Second, much of the comparative-effectiveness literature evaluating disinfectants relies on observational single-site studies, and early validation frameworks often depended on self-reported survey data; randomised comparison in occupied healthcare buildings is frequently impractical, though recent systematic reviews now supply quantitative synthesis earlier work lacked. Third, regulatory thresholds remain expressed in the units of the method whose deficiencies motivated the alternatives, and the alignment of verification standards with advances in molecular detection remains an ongoing challenge. Full article

Low adult density can suppress recruitment through mate limitation or cooperative reproductive failure, even when a host improves adult survival. Motivated by this ecological tension, we introduce a fecundity-related component Allee effect into a stage-structured commensalism model with a Holling type II host benefit. The nonlinear recruitment term is $f\left(x_2\right)=\alpha x_2^2/(x_2+A)$, and the independently growing host converges to its carrying capacity, yielding an asymptotically autonomous commensal subsystem. We obtain a complete analytical classification. If the long-term host benefit D exceeds mature mortality ${\delta }_2$, a unique coexistence equilibrium attracts every nontrivial commensal population. If the benefit is insufficient, fecundity limitation can instead create two positive equilibria: a low-density saddle whose stable set is the basin boundary and a stable high-density coexistence state. Thus, extinction and coexistence are both possible, depending on initial stage composition and host abundance. Stronger fecundity limitation enlarges the extinction region, whereas greater recruitment, maturation or host benefit promotes persistence. At the critical boundary, the positive equilibria merge in a non-degenerate saddle-node bifurcation. Rigorous global arguments and numerical comparisons with the linear-recruitment model show how a component Allee effect can induce strong-Allee-type demographic behavior, including bistability and threshold-mediated extinction. Full article

Background/Objectives: Food exclusions in vegan and vegetarian diets may complicate the assessment of disordered eating and orthorexia. This study compared eating disorder risk, healthy orthorexia, and orthorexia nervosa across vegan, vegetarian, and omnivorous women and examined their adjusted associations with dietary pattern. Methods: This cross-sectional study used convenience sampling to recruit 305 women in Türkiye who self-identified as vegan (n = 64), vegetarian (n = 71), or omnivorous (n = 170). Participants completed the Eating Attitudes Test-26, Teruel Orthorexia Scale, and Food Choice Questionnaire. Associations with vegan and vegetarian patterns were examined using multinomial logistic regression, with omnivorous women as the reference group. Potential nonlinearity in the association between orthorexia nervosa and the vegan versus omnivorous contrast was examined using restricted cubic splines. Secondary exploratory food choice models were corrected for multiple testing. Results: Elevated eating disorder risk was not significantly associated with either the vegan or vegetarian pattern in the adjusted model. Higher healthy orthorexia scores were associated with greater odds of following a vegan pattern (adjusted odds ratio (aOR) = 1.177, 95% confidence interval (CI): 1.090–1.270), whereas the association between orthorexia nervosa and the vegan pattern was nonlinear (p for nonlinearity = 0.006), with the adjusted probability declining across low-to-moderate scores and the estimates becoming imprecise at higher scores. Neither orthorexia dimension was associated with the vegetarian pattern. In secondary exploratory analyses, weight control motivation was associated with lower odds of vegan (aOR = 0.424, 95% CI: 0.249–0.723) and vegetarian patterns (aOR = 0.610, 95% CI: 0.382–0.975), while familiarity was associated with lower odds of the vegan pattern (aOR = 0.466, 95% CI: 0.296–0.735). Conclusions: Healthy orthorexia and orthorexia nervosa showed distinct associations with the vegan pattern. Dietitians and healthcare professionals should not infer eating pathology from dietary pattern labels alone but should consider motivations for food exclusions, weight- and shape-related concerns, dietary rigidity, distress, and functional impairment. Full article

The tympanic membrane (TM) is a thin, semi-transparent structure whose morphology and optical appearance provide important diagnostic cues. In the early stages of middle-ear pathology, subtle shape and compliance alterations may precede overt clinical signs, making them valuable early indicators of disease. Such structural changes are difficult to assess reliably using conventional (micro-)otoscopy, which lacks quantitative depth information and is operator-dependent. Data-driven monocular image analysis could enable quantitative assessment of TM geometry and compliance, but the limited availability of annotated three-dimensional datasets constrains the development of these methods. At the same time, realistic simulations of TM appearance remain challenging due to its complex reflectance and transmission behavior. The present study focuses on physiologically healthy tympanic membranes, which provide the baseline anatomical and optical model required before subtle pathological changes can be investigated. This work introduces a parametric physically grounded rendering model of the structures visible during otoscopy: the tympanic membrane, ear canal, and malleus–incus complex. Implemented in the open-source software Blender™ using procedural geometry nodes and physically motivated shaders, the framework generates anatomically plausible three-dimensional geometries via statistical parameter sampling and controlled mesh deformation. Optical appearance is simulated using a computationally efficient layered shading model based on literature-derived tissue reflectance, transmission, and scattering properties. A camera–projector setup models both conventional white-light otoscopy and structured-light imaging, enabling the generation of paired intensity images and corresponding depth maps. The proposed framework establishes a physically grounded representation of the human ear and enables a controllable, extensible modeling pipeline for virtual training, biomechanical finite element analysis, and synthetic data generation for supervised learning. Full article

Adaptive learning environments and large language models (LLMs) are increasingly combined, yet most LLM-based tutoring work addresses feedback generation in isolation rather than its integration into a controlled, auditable learning loop. This paper presents a design study of a provider-independent LLM architecture for adaptive Oracle SQL tutoring. We report the system at the design stage—specifying the underlying adaptive learning platform, a provider-independent service layer that abstracts interchangeable model back-ends behind a single auditable interface, the SQL-specific extension (a reference-based evaluator demonstrated as a proof-of-concept on an in-memory SQLite fixture of the fixed Oracle HR schema; Oracle-specific execution, the live sandbox, and course-wide evaluation remain planned), and a concrete validation pathway—without claiming empirical effectiveness. The design pairs cognitive adaptivity, realised as mastery-based task selection under configurable parameters, with motivational design grounded in Keller’s ARCS model. Following a design-based research orientation, the adaptation logic is expressed as testable design conjectures. We contribute (a) a modular, provider-independent architecture for LLM-supported SQL learning; (b) a formalised adaptation logic linking learner states, adaptive responses, and conjectures; (c) an illustrative prompt template with a three-level quality-control workflow for AI-generated feedback; and (d) a proposed estimation-oriented feasibility pilot protocol—a validation pathway rather than an empirical contribution of this article—to be preregistered before data collection, with up to 50 undergraduate participants. Implemented capabilities, planned SQL-specific functions, and hypothesised outcomes are kept explicitly distinct throughout. Full article

Background: This study aimed to investigate the prevalence and levels of depressive, anxiety, and stress symptoms among dentists at an academic institution, and to identify demographic and work-related factors correlated with these psychological conditions. Methods: A cross-sectional study was conducted among full-time dentists teaching at King Saud University (Riyadh, Saudi Arabia). Data were collected using a self-administered questionnaire that captured demographic and work-related variables, alongside the validated 21-item Depression, Anxiety, and Stress Scale (DASS-21). Data were analyzed using descriptive statistics, Chi-square tests, and multiple linear regression to determine significant associations. Results: A total of 114 participants completed the questionnaire, of which 86.8% were faculty and the rest non-faculty. The prevalence of depressive, anxiety, and stress symptoms was 21.9%, 32.5%, and 34.2%, respectively, predominantly at moderate levels. The prevalence of at least one subscale was 44.7%, and a significant correlation was found for the co-occurrence of the three subscales (p < 0.001). Regression analysis indicated three correlates of stress: dissatisfaction, female sex, and choosing a career in academia were significantly associated with the stress scores (p < 0.05). No significant factors were correlated with the anxiety or depression subscales. While females and dissatisfied participants reported higher scores for stress, intending to work in academia as a primary career choice was significantly correlated with lower levels of stress, highlighting the potential importance of intrinsic motivation for academia. Conclusions: A considerable proportion of participants experience depression, anxiety, and stress symptoms, suggesting the need for targeted institutional strategies to support mental well-being, particularly among vulnerable groups. Full article