Search Results (6,448)

Background: Post-COVID-19 condition (PCC) affects many survivors, with evidence of sex-specific differences in prevalence and symptom profiles. However, few prediction studies have examined whether sex-stratified models improve prediction or generalize across sexes. This study aimed primarily to develop and compare sex-stratified machine learning models for PCC prediction using routinely available baseline variables, and secondarily to assess cross-sex generalizability and adversarial robustness. Methods: We analyzed a prospective longitudinal cohort of 1006 adults hospitalized with COVID-19 at Sechenov University Hospital Network (Moscow, Russia). Demographics, smoking status, and pre-existing comorbidities were extracted from medical records, and PCC status was assessed at 6-month follow-up. Machine learning models—including classical algorithms and graph-based neural networks—were trained separately for males and females. Cross-sex validation evaluated generalizability, variable importance aided interpretation, and adversarial perturbations assessed model robustness. Results: PCC prevalence was higher in females (53.9%) than males (39.1%). Overall predictive performance was modest across all models, with AUC values ranging approximately 0.50–0.61. Graph-based models achieved the highest discrimination, with the best AUC reaching approximately 0.61, while classical approaches provided limited predictive value. Cross-sex validation showed minor asymmetry: models trained on male data performed slightly better on female cases than vice versa. Adversarial testing revealed sensitivity of all models to input perturbations. Conclusions: Demographics and comorbidities alone provide insufficient information for reliable PCC prediction. Modest sex-specific differences in model generalizability suggest distinct, sex-associated PCC phenotypes, but richer multimodal data—including clinical biomarkers, wearable-derived measures, and patient-reported outcomes—will be required to develop clinically useful and equitable predictive models. Sex-stratified approaches should be considered in future post-viral syndrome prediction studies. Full article

Purpose: In this study, our aim is to assess the efficacy of saline-assisted needle withdrawal in minimizing pneumothorax and bleeding complications during CT-guided lung biopsies. Methods: In this retrospective study, 400 patients who underwent CT-guided lung biopsy were divided into two groups: 200 patients underwent conventional needle withdrawal (control group), and the other 200 patients underwent saline-assisted needle withdrawal (NaCl group). Needle angle, patient positioning, demographic data, and other procedural variables were collected. The primary outcome was the incidence of pneumothorax; secondary outcomes included bleeding rates and the impact of procedural factors on complication risk. Statistical analyses included Chi-square tests, logistic regression, and multinomial modeling. Results: The NaCl group demonstrated a significantly lower incidence of clinically significant pneumothorax (13.5%) compared to the control group (22.0%) (p = 0.007), while bleeding complications occurred in 26.5% of patients in the former versus 45.0% in the latter (p < 0.001). Multivariate analysis suggested a non-significant trend toward a higher pneumothorax risk with shallow needle angles (<60°; p = 0.502). Saline injection was especially advantageous in patients with underlying lung disease, reducing pneumothorax severity even when overall incidence rates were similar. No adverse events were attributed to the use of saline. Conclusions: Saline-assisted tract sealing is a safe and cost-effective technique that significantly reduces the risk of clinically significant pneumothorax and bleeding in CT-guided lung biopsies. Due to its simplicity and favorable safety profile, this approach holds considerable promise for widespread clinical implementation. Full article

Outdoor music festivals (OMFs) increasingly operate as a form of temporary land use that activates urban stock land, yet how such land-use reconfigurations unfold across uneven urban–rural geographies remains insufficiently examined. Taking China’s OMFs from 2013 to 2024 as a case, this study applies the Geodetector model within a spatial justice framework to analyze fifteen indicators organized along the distributional, procedural, and recognition dimensions. The results show a pronounced “market-sinking” trend accompanied by westward expansion, and the seasonal clustering gradually moderated. The three dimensions jointly shape OMFs’ diffusion through distinguishable pathways, with the procedural dimension exhibiting the highest explanatory power through institutional steering and industrial coordination, followed by the recognition dimension through demographic foundations and digital visibility, and the distributional dimension through material and infrastructural accessibility; interaction detection further indicates that their joint presence produces amplified effects. These mechanisms align with international land-use and territorial-governance studies, while reflecting the state-led coordination distinctive to China. The findings point to an emerging form of collaborative co-creation in which governmental, market, and community actors jointly shape the spatial production of cultural events, extending the classical core–periphery account and informing debates on the territorial governance of event spaces in non-metropolitan regions. Full article

Objective: This study aimed to identify clinically modifiable and readily accessible predictors of 30-day mortality in a 10-year candidemia cohort and to assess temporal changes in Candida species distribution. Methods: We retrospectively evaluated 391 hospitalized adults with positive blood cultures for Candida spp. between January 2015 and March 2025. Only the first candidemia episode was included. Demographic characteristics, comorbidities, risk factors, laboratory parameters, antifungal therapy, and outcomes were recorded. Species identification was performed using conventional methods and the VITEK 2 system. Factors associated with 30-day mortality were analyzed using univariate and multivariate logistic regression models. Results: The mean age was 64.5 ± 17.7 years, and 56.3% of patients were male. Most patients (68.8%) were managed in the intensive care unit, and the 30-day mortality rate was 54%. Non-albicans Candida species accounted for 62.7% of isolates, with an increasing trend over time, particularly for Candida glabrata. Fluconazole susceptibility was 79%. In univariate analysis, advanced age, solid tumors, invasive mechanical ventilation, leukocytosis, thrombocytopenia, septic shock, intensive care unit admission, and failure to remove the central venous catheter were associated with mortality. Multivariate analysis identified advanced age, intensive care unit admission, septic shock, failure to remove the central venous catheter, leukocytosis, and thrombocytopenia as independent predictors of 30-day mortality. Conclusions: Candidemia remains a life-threatening infection with high mortality. Central venous catheter management and simple hematological parameters, particularly white blood cell and platelet counts, provide practical tools for early risk stratification. Although the rising prevalence of non-albicans Candida species may require updates in empirical therapy, prompt source control remains essential to improve survival. Full article

Introduction: The relationship between epilepsy and alcohol use is complex and clinically significant. Alcohol acts as a neurochemical modulator capable of lowering the seizure threshold during both intoxication and withdrawal, while chronic misuse may contribute to epileptogenesis through neuronal injury, metabolic stress, and neurotransmitter dysregulation. However, the long-term impact of alcohol use on mortality among people with epilepsy (PWE) remains insufficiently characterized. The purpose of this study was to assess all-cause mortality risk among individuals with epilepsy based on alcohol use history, stratified by race/ethnicity. Methods: Data from the 2008–2018 National Health Interview Survey (NHIS) were linked to mortality outcomes on 31 December 2019 from the National Death Index (NDI) for U.S. adults aged 18 years and older. PWE and alcohol use were determined using self-reported data. Survival probabilities were estimated using weighted Kaplan–Meier methods, and hazard ratios were calculated using Cox proportional hazards models adjusted for demographic and clinical covariates. Results: Our results indicated that among PWE, alcohol use was associated with increased all-cause mortality. The unadjusted hazard ratio (HR) for alcohol use among individuals with epilepsy was 1.30, increasing to 1.40 after multivariable adjustment. In contrast, alcohol use alone without epilepsy was not associated with elevated mortality risk after adjustment. When stratified by race, the combined effect of epilepsy and alcohol use was significantly associated with increased mortality among Black individuals but not White individuals. Conclusions: In this nationally representative cohort, the combined presence of epilepsy and alcohol use was associated with higher all-cause mortality compared with alcohol use alone. Racial differences were observed, underscoring the need for integrated clinical care and further research into genetic, biological, and social determinants influencing epilepsy outcomes. Full article

Objective: To examine the associations of metabolic health and obesity phenotypes with liver fat accumulation and hepatic steatosis in adults. Methods: We analyzed 676 non-Hispanic white adults (18–95 years; 55.8% female) from the Fels Longitudinal Study using a cross-sectional design. Participants were classified into metabolically healthy normal weight (MHNW), metabolically healthy obesity (MHO), metabolically unhealthy normal weight (MUNW), and metabolically unhealthy obesity (MUO) phenotypes. Metabolically unhealthy status was defined as the presence of ≥1 metabolic dysfunction, consistent with prior epidemiological definitions; secondary analyses using ≥2 were also performed. Obesity was defined using DXA-derived body fat percentage. Liver fat (%) was quantified using magnetic resonance imaging, and hepatic steatosis was defined as liver fat > 5.56%. Multivariable linear and probit regression models were used to evaluate associations, adjusting for demographic and lifestyle covariates; secondary models additionally examined dietary intake. Results: Mean liver fat was 5.95% (SE = 0.23), and steatosis was present in 29.8% of participants. Compared to MHNW individuals, liver fat was significantly higher in MHO (mean 3.77% vs. 2.67%), MUNW (4.63%), and MUO (8.47%) phenotypes. After covariate adjustment, liver fat was 33.8% (95% CI: 13.7–57.5%) higher in MHO, 28.1% (10.1–49.0%) higher in MUNW, and 113.0% (85.3–144.7%) higher in MUO relative to MHNW. Corresponding increases in steatosis probabilities were observed across phenotypes. No individual dietary component or dietary pattern was significantly associated with liver fat after adjustment. Conclusions: Metabolically healthy obesity was associated with higher liver fat and steatosis probability compared with metabolically healthy normal weight, with levels comparable to metabolically unhealthy normal weight individuals. These findings suggest that the absence of overt metabolic abnormalities does not necessarily indicate a metabolically benign state with respect to liver fat accumulation. Given the cross-sectional design, these results should be interpreted as associations rather than causal relationships. Full article

Obesity represents a multifactorial health condition influenced by complex interactions among behavioral, environmental, and physiological factors, yet the relative predictive importance of lifestyle behaviors versus medical history indicators remains incompletely characterized. This investigation employed a three-phase machine learning approach to systematically compare the predictive power of behavioral lifestyle factors, medical history variables, and their integration for obesity classification. Phase A utilized a dedicated obesity dataset containing demographic, dietary, and lifestyle predictors to perform seven-category obesity classification, achieving 81.65% test accuracy with an optimized Random Forest ensemble and macro-averaged F1-score of 0.82. Phase B addressed binary obesity classification using health indicators from diabetes screening data, where a Gradient Boosting model with optimized decision threshold achieved 67.84% accuracy and AUC of 0.735, demonstrating substantially lower performance than behavioral predictors. Phase C integrated both feature sets into a unified model, where Gradient Boosting achieved 68.31% accuracy and AUC of 0.747, representing marginal improvement over medical history alone. Cross-validated performance comparisons revealed that behavioral lifestyle factors provided superior discriminative power compared to medical history indicators, with dedicated lifestyle predictors achieving 13.81 percentage points higher accuracy than medical indicators. Feature importance analysis confirmed that transportation mode, physical activity patterns, and dietary behaviors ranked among the most influential predictors in the combined model. These findings demonstrate that behavioral lifestyle factors constitute stronger obesity predictors than medical history variables, with implications for clinical screening strategies and public health intervention targeting that prioritize lifestyle assessment and modification programs. Full article

The COVID-19 pandemic has renewed interest in proximity tourism as a driver for culturally sustainable local development. Against this backdrop, this study examines the Province of Latina, Italy, evaluating the potential of proximity tourism flows originating from the Metropolitan City of Rome to reduce the socioeconomic and infrastructural disparities that affect the provincial territory. Using a multi-stage quantitative framework, the research identifies key determinants of municipal attractiveness and assesses the structural relationships between tourism activities, economic development, and demographic trends. Results reveal a clear spatial polarization: while coastal municipalities successfully leverage natural and cultural assets for development, they increasingly face challenges related to overtourism and gentrification. Conversely, marginalized hilly and mountainous peripheries remain underdeveloped despite their high heritage value. Empirical evidence suggests that while proximity tourism does not directly drive demographic growth, it acts indirectly as a catalyst by stimulating economic development. The study concludes that proximity tourism can foster territorial cohesion and socioeconomic resilience only when coupled within integrated strategies that improve accessibility, valorize peripheral assets, and keep the added value of tourism embedded within host communities. These findings offer actionable insights for policymakers aiming to transition toward more inclusive and culturally sustainable tourism models in the post-pandemic era. Full article

Chest radiography remains a cornerstone in the diagnosis of thoracic diseases. However, differences in image acquisition—particularly projection type—may influence the apparent prevalence and detectability of radiographic findings. Such differences may represent a potential source of bias in large imaging datasets used for clinical research and artificial intelligence. Importantly, projection type is closely associated with the patient’s condition and may therefore reflect both technical imaging factors and underlying clinical characteristics, including disease severity. A total of 120,120 chest radiographs were available in the dataset. After applying inclusion criteria, 112,104 images were included in the primary analysis. Multivariable logistic regression models were used to assess the association between projection type and the presence of radiographic findings, adjusted for age and sex. Subgroup and interaction analyses were performed to evaluate effect modification by demographic factors. Given the large sample size, emphasis was placed on effect sizes and confidence intervals rather than statistical significance alone. Compared with posteroanterior projection, anteroposterior projection was associated with higher odds of detecting consolidation (aOR 3.27; 95% CI 3.07–3.48), infiltration (aOR 1.90; 95% CI 1.84–1.96), pleural effusion (aOR 1.66; 95% CI 1.60–1.72), atelectasis (aOR 1.63; 95% CI 1.57–1.70), and cardiomegaly (aOR 1.19; 95% CI 1.10–1.28). These associations were consistent across age and sex strata. A significant interaction between projection type and sex was observed for infiltration (p = 0.01). Projection type is associated with substantial differences in the detection of thoracic abnormalities on chest radiographs. These associations should be interpreted with caution, as they likely reflect a combination of technical imaging effects and residual confounding related to patient severity and clinical context. Projection may therefore act as a marker of dataset heterogeneity rather than a purely causal factor. Accounting for projection metadata is therefore essential to improve clinical interpretation and to ensure the robust development and validation of artificial intelligence models. Full article

Background/Objectives: A floating hip injury, defined as an ipsilateral fracture of the pelvis or acetabulum combined with a femoral fracture, represents a rare and devastating musculoskeletal injury resulting from high-energy trauma. Although Müller type A floating hip injuries comprising an acetabular fracture with an ipsilateral femoral fracture are recognized for their clinical complexity, the long-term prognostic factors influencing functional outcomes remain poorly elucidated. This study aimed to identify independent prognostic factors associated with unsatisfactory long-term functional outcomes in patients with Müller type A floating hip injuries. Methods: A retrospective study was performed on 68 consecutive patients with Müller type A floating hip injuries who underwent surgical fixation at a single tertiary trauma center, with a minimum follow-up period of 5 years. Functional outcomes were assessed using the Majeed score, and patients were dichotomized into satisfactory (n = 48; 70.6%) and unsatisfactory (n = 20; 29.4%) outcome groups. Acetabular fractures were classified according to the Judet–Letournel system, and femoral fractures were classified by fracture level (proximal, shaft, or distal). Radiological outcomes were evaluated using Matta’s radiological grading system. Demographic, injury-specific, and treatment-related variables were compared between groups using the Mann–Whitney U test and chi-square test with Bonferroni correction. A multivariate binary logistic regression model was constructed to determine independent predictors of unsatisfactory outcomes. Results: The mean age was 37.15 ± 12.07 years, with a male predominance (67.6%). The predominant mechanism of injury was pedestrian struck by vehicle (54.4%), followed by motor vehicle collision (27.9%) and fall from height (17.6%); collectively, high-energy vehicular trauma accounted for 82.3% of cases. In the univariate analysis, transverse with posterior wall acetabular fracture pattern (p = 0.001), proximal femur fracture level (p = 0.001), associated lower extremity fractures (p = 0.001), nerve damage (p = 0.001), higher body mass index (BMI) (p = 0.001), and lower Matta’s radiological scores (p = 0.001) were significantly associated with unsatisfactory outcomes. Three independent predictors emerged in the multivariate logistic regression: BMI (OR = 1.50; 95% CI: 1.05–2.15; p = 0.025), the presence of associated lower extremity fractures (OR = 29.02; 95% CI: 2.83–297.67; p = 0.005), and Matta’s radiological score (OR = 0.06; 95% CI: 0.01–0.56; p = 0.014). The model yielded internal discriminatory metrics within the acceptable range (overall accuracy 89.7%, sensitivity 95.8%, specificity 75.0%, Nagelkerke R² = 0.757); however, given the limited events-per-variable ratio (~6.7) and the wide confidence intervals observed for some predictors, these internal performance estimates are likely optimistic due to potential overfitting, and the findings should be interpreted as exploratory pending external validation. Conclusions: Elevated BMI, the presence of associated ipsilateral lower extremity fractures, and poor quality of acetabular reduction, assessed via Matta’s radiological criteria, are independent determinants of unsatisfactory long-term functional outcomes in Müller type A floating hip injuries. These findings underscore the critical importance of achieving anatomical reduction in the acetabulum and highlight the compounding effect of additional ipsilateral limb injuries on patient prognosis. Full article

Tibetan pigs are an important indigenous genetic resource on the Qinghai-Tibetan Plateau, but their population history remains unclear. Here, we analyzed whole-genome resequencing data from 29 Tibetan pigs and 69 wild boars from different regions to investigate their genetic structure and origin. Population structure analyses showed that Tibetan pigs formed a distinct genetic cluster, but clear substructure was present among the four Tibetan pig groups. Demographic analyses indicated that all Tibetan pigs shared a similar deep ancestral background, whereas the Tibetan population from Tibet (TT) showed a distinct recent demographic trajectory relative to the other Tibetan pig groups. Model inference based on multidimensional site frequency spectra further supported a mixed origin of TT from northern and southern Asian wild boar-related ancestral components. In contrast, the other three Tibetan pig groups were best explained as lineages derived from TT after its formation. We also detected a persistent western Eurasian-related excess-sharing signal in TT, consistent with potential introgression. Selection scans identified candidate genes and pathways potentially related to hypoxia adaptation, cardiovascular function, and lung development. Full article

Background/Objectives: The prevalence of allergic diseases has markedly increased in developed countries, with environmental and dietary factors considered important contributors. Folic acid is an essential micronutrient involved in one-carbon metabolism and DNA methylation, playing a key role in epigenetic regulation of immune function. Both high and low folate exposure have been associated with allergic outcomes, but the data on postnatal folate status in paediatric populations remain limited. This study aimed at assessing serum folate status in children with atopic diseases compared with children without chronic allergic disease in Croatia. Methods: This cross-sectional study included 292 paediatric patients from the University Hospital in Split and a paediatric primary care practice between January 2024 and January 2025. Serum folic acid concentrations were measured using electrochemiluminescence immunoassay. Additional laboratory parameters included vitamin B12, total IgE levels, and eosinophil counts. Demographic and clinical data were obtained from medical records. Statistical analyses included Chi-square tests, Mann–Whitney U tests, linear regression modelling, and analysis of covariance with statistical significance set at p < 0.05. Results: Folic acid deficiency was present in 66.4% of all participants. Children with atopic diseases were significantly more likely to have folate deficiency and had lower mean serum folate concentrations compared to children without allergic disease. There were no significant differences in folate levels between children with and without asthma. Lower folate levels were associated with higher IgE levels, higher eosinophil counts, and older age. When controlling for the effects of age on folic acid levels, the differences between participants with and without atopic diseases remained significant. Conclusions: Folic acid deficiency is highly prevalent among children in the Mediterranean region of Croatia and is significantly associated with atopic diseases and markers of allergic inflammation. These findings highlight a potential role of folate status in paediatric allergic disease and support the need for longitudinal studies to clarify causality and potential clinical implications. Full article

Background: Varicella burden is closely linked to national socioeconomic development, yet systematic analyses of its non-linear relationship with the Socio-demographic Index (SDI) are lacking. This study aims to elucidate this relationship and inform equitable, context-specific strategies. Methods: Based on data from the Global Burden of Diseases 2021 study, we analyzed global trends (1990–2021) in the incidence, prevalence, mortality, and disability-adjusted life-years (DALYs) of varicella. Joinpoint regression was used to identify trend transition points, and an autoregressive integrated moving average (ARIMA) model was applied to forecast the disease burden through 2035. Analyses were conducted, and countries and territories were stratified into five SDI groups: high (SDI > 0.81), high–middle (0.70–0.81), middle (0.61–0.69), low–middle (0.46–0.60), and low (SDI < 0.46). These approaches aimed to systematically elucidate the socioeconomic gradient of the varicella burden and to specifically investigate its potential non-linear relationship with SDI. Results: From 1990 to 2021, global age-standardized mortality and DALYs declined by −45.71% (95% UI: −48.32% to −42.95%) and −36.15% (95% UI: −39.04% to −33.01%), respectively, while incidence and prevalence rates slightly increased. A significant U-shaped relationship emerged between burden and SDI, with rates highest in low- and high-SDI regions. The rise in high-SDI regions was driven by increasing incidence and prevalence from 1996 to 2015. Projections to 2035 indicate continued global decline but persistent disparities. Conclusions: The varicella burden follows a U-shaped socioeconomic gradient. Rising incidence in high-SDI regions highlights that economic development and routine pediatric vaccination alone are insufficient. Precision strategies tailored to SDI levels—closing adult immunity gaps in high-SDI, sustaining gains in middle-SDI, and expanding vaccine access in low-SDI regions—are essential. Full article

Pancreatic cancer is a leading cause of cancer-related mortality in the United States and worldwide. Most patients are diagnosed with pancreatic cancer at advanced stages, when curative therapy is no longer possible. The stage of pancreatic cancer at diagnosis critically impacts the treatment options and thus the clinical outcomes. Currently, there is no established screening program or tests for its early detection. Studying and understanding how those factors influence the stage of pancreatic cancer at diagnosis helps identify barriers and develop screening strategies. Tumoral and demographic factors, as well as social determinants of health, tend to be associated with localized vs. advanced stage of pancreatic cancer at diagnosis. Socioeconomic factors have been shown to be important mediators of racial disparities in stage at diagnosis as well as germline genetic testing. Recently, screening initiatives, blood-based molecular biomarker tests for early detection of pancreatic cancer, and machine learning-based models for risk prediction and imaging diagnostics have been developed. By determining and understanding the factors associated with the stage at diagnosis, risk-stratified screening can be feasible by combining demographics, genetics, comorbidities, lifestyle, and social determinants. Moreover, regulatory policies that address the social determinants of health can guide the development of screening strategies to allocate resources for equitable access to healthcare and to reduce disparities in patients with pancreatic cancer. Full article

Introduction: Colorectal cancer remains a leading cause of cancer-related morbidity and mortality, with adenomatous polyps representing a common precursor. Post-polypectomy polyp recurrence represents a significant risk of colorectal cancer, driving periodic colonoscopy surveillance and polypectomy as needed. In this study, we explore a multimodal machine learning approach that integrates endoscopic imaging with clinical and pathology data to improve recurrence risk prediction and support individualized surveillance planning. Methods: We developed and evaluated a multimodal artificial intelligence (AI) model to predict post-polypectomy colorectal polyp recurrence using the ERCPMP-v5 dataset. The cohort included 217 patients with 796 high-resolution endoscopic RGB images and 21 endoscopic videos; video data were converted to still frames at 2 frames per second. Images and frames were resized to 224 × 224 pixels and normalized. Patient-level demographic, morphological (Paris, Kudo Pit, JNET), anatomical, and pathological variables were encoded using standard scaling for continuous features and one-hot encoding for categorical features. Visual representations were extracted using a pretrained Vision Transformer backbone (ViT-Base-Patch16-224) with frozen weights. Structured metadata (79 variables) was encoded using a multilayer perceptron. A late fusion framework used image and metadata representations to generate a recurrence probability via a sigmoid classifier; probabilities were thresholded at 0.5 for binary prediction. Model performance was evaluated on a held-out test set using accuracy, precision, recall, F1-score, and area under the receiver operating characteristic curve (AUC). We additionally compared fusion performance with image-only and metadata-only baselines. Predicted probabilities were translated to surveillance recommendations using risk tiers: low risk (0.00 ≤ p < 0.20), moderate risk (0.20 ≤ p < 0.50), and high risk (p ≥ 0.50). Results: On the test set, the multimodal fusion model achieved 90.4% accuracy, 86.7% precision, 83.1% recall, 84.9% F1-score, and an AUC of 0.920. The image-only model achieved 84.6% accuracy (AUC 0.880), and the metadata-only model achieved 81.9% accuracy (AUC 0.850), indicating improved performance with multimodal fusion. Risk stratification enabled surveillance recommendations of 1–3 years for low risk, 6–12 months for moderate risk, and 3–6 months for high risk. Conclusions: A late-fusion multimodal model integrating endoscopic imaging with structured clinical and pathology variables demonstrated excellent performance for predicting post-polypectomy recurrence and generated actionable risk-based surveillance intervals. This approach may support individualized follow-up planning and more efficient allocation of surveillance resources, while prioritizing timely evaluation for patients at higher predicted risk. Full article