Tag: statistics

Front Nephrol. 2026 Jul 27;6:1818921. doi: 10.3389/fneph.2026.1818921. eCollection 2026.

ABSTRACT

BACKGROUND: Infection is the second leading cause of death and hospitalization in hemodialysis (HD) patients. The occurrence of infections in HIV/AIDS patients with End-Stage Renal Disease (ESRD) undergoing maintenance hemodialysis is influenced by patient characteristics and related treatment regimens. Currently, there is limited data both domestically and internationally on the risk factor analysis for adverse outcome following infections in HIV/AIDS patients with ESRD receiving maintenance hemodialysis. This study aims to identify the risk factors associated with adverse outcomes following infections during hemodialysis in patients with HIV/AIDS and ESRD.

METHODS: In this retrospective cohort study, we evaluated the 9-month treatment outcomes of 70 HIV/AIDS patients with ESRD and concurrent infections undergoing maintenance hemodialysis at a hospital in Guangxi between January 2019 and December 2024. The patients were divided into adverse outcome group(defined as mortality or severe complications leading to prolonged hospitalization) and survival group. Data analysis was performed using IBM SPSS Statistics version 26.0, with statistical significance set at P ≤ 0.05, to identify adverse outcome risk factors for concurrent infections in HIV/AIDS patients with ESRD.

RESULTS: Among the 70 enrolled patients, the age ranged from 11 to 86 years. Adverse outcomes occurred in 48 cases (68.57%), while 22 patients (31.43%) survived without severe events. Multivariate analysis identified low-level hemoglobin(HB)(OR 0.952(95%CI 0.908,0.998),P = 0.042) and platelet (PLT)(OR 0.987(95%CI 0.975,1.000),P = 0.046).

CONCLUSION: Adverse outcomes are common among HIV/AIDS patients with ESRD who develop infections during maintenance hemodialysis. Low hemoglobin and low platelet levels are significant risk factors for poor prognosis in this patient population. Early identification and management of these hematologic abnormalities may help improve clinical outcomes.

PMID:42577440 | PMC:PMC13454224 | DOI:10.3389/fneph.2026.1818921

Front Immunol. 2026 Jul 27;17:1870438. doi: 10.3389/fimmu.2026.1870438. eCollection 2026.

ABSTRACT

BACKGROUND: Latent autoimmune diabetes in adults (LADA) is characterized by progressive β-cell impairment and severe glycemic lability, predisposing patients to in-hospital hypoglycemia. Few tailored risk-stratification models exist for this population. This study aimed to develop and validate an interpretable machine learning model using routine clinical data to predict in-hospital hypoglycemia in LADA inpatients.

METHODS: This multicenter retrospective study recruited participants from five Chinese tertiary hospitals between January 2019 and September 2025. Data from four centers formed the derivation cohort, and the remaining center served as the independent external validation cohort. The primary endpoint was in-hospital hypoglycemia (blood glucose < 3.9 mmol/L). Three machine learning models, including logistic regression, random forest, and XGBoost, were developed using routine clinical data and assessed for discrimination, calibration, and clinical utility. SHAP analysis was applied to improve model interpretability. Exploratory subgroup analyses in the internal validation cohort examined model performance across clinical subgroups.

RESULTS: A total of 752 LADA inpatients were enrolled. The incidence of in-hospital hypoglycemia was 44.8% in the derivation cohort and 54.4% in the external validation cohort. Six core predictive factors were identified: largest amplitude of glycemic excursion, fasting C-peptide, glycated hemoglobin, sex, insulin pump use, and previous hypoglycemia. The three models yielded numerically variable discriminative performance across cohorts. Pairwise DeLong tests indicated no statistically significant differences in the AUROC among the three algorithms during external validation. All models showed comparable calibration and threshold-dependent predictive performance in the external cohort. XGBoost was selected as the final model after comprehensive evaluation. Fasting C-peptide was identified as the most influential predictor. Exploratory subgroup analyses demonstrated generally stable model performance across clinical strata. These findings are limited by small subgroup sample sizes and wide confidence intervals, and thus cannot be generalized to external populations. Sensitivity analysis suggested that model performance was not predominantly dependent on the retained glucose-derived predictor.

CONCLUSIONS: The interpretable XGBoost model showed acceptable discrimination, calibration, and potential clinical utility for in-hospital hypoglycemia risk stratification in patients with LADA. This pragmatic predictive tool has the potential to support individualized inpatient glycemic management and facilitate targeted clinical intervention for LADA populations.

PMID:42577431 | PMC:PMC13454309 | DOI:10.3389/fimmu.2026.1870438

Health Sci Rep. 2026 Aug 9;9(8):e73029. doi: 10.1002/hsr2.73029. eCollection 2026 Aug.

ABSTRACT

BACKGROUND AND AIMS: Armed conflict in Sudan has severely disrupted traditional medical education, prompting a shift toward online learning as an alternative. While digital platforms enable educational continuity, they are associated with increased screen exposure and a rising incidence of digital eye strain (DES) among medical students. This study aimed to investigate the prevalence, symptoms, and associated factors of DES in this conflict-driven educational environment. To assess the prevalence, risk factors, and awareness of DES symptoms among Sudanese medical students engaged in online learning during the armed conflict.

METHODS: A descriptive cross-sectional study was conducted from April to June 2025 among undergraduate Sudanese medical students who transitioned to online learning during the conflict. A total of 1028 participants were recruited using convenience sampling. Data were collected using an online self-administered questionnaire adapted from a validated instrument used in a previously published study. Responses were analyzed using SPSS version 27. Descriptive statistics were used to summarize findings, binomial logistic regression was performed to identify predictors of DES symptoms, and the Wilcoxon signed-rank test assessed changes in screen time before and during the conflict.

RESULTS: About 80.6% of students reported that prolonged screen use negatively impacted their eye health and lifestyle. Eye-related symptoms were reported by 86.6% of participants, and 61.5% experienced physical discomfort. The average screen time increased significantly during the conflict. Awareness of the 20-20-20 rule was limited (24.2%), with only 13.5% reporting regular practice. However, 70.9% were willing to reduce screen time to help prevent DES.

CONCLUSION: Digital eye strain was highly prevalent among medical students and was associated with extended screen use, poor posture, and limited awareness. Its association with physical discomfort and academic stress highlights the need for targeted interventions. Medical institutions should integrate DES awareness and prevention strategies into curricula to promote student well-being and academic success.

PMID:42577426 | PMC:PMC13454369 | DOI:10.1002/hsr2.73029

Front Bioinform. 2026 Jul 27;6:1851023. doi: 10.3389/fbinf.2026.1851023. eCollection 2026.

ABSTRACT

Large language model (LLM) text embeddings have recently been used for supervised learning on tabular data by serializing each observation into text and then converting the text into a dense fixed-length vector. This strategy is appealing for biomedical tabular data, which are often mixed-type and contain missing values, because it produces a complete numeric representation even when some original entries are missing. However, its suitability for unsupervised tasks remains unclear. Here we evaluate the use of LLM-derived text embeddings for dimension reduction of tabular data, focusing on biological and clinical datasets. We compare an LLM embedding-based approach with a direct tabular approach that computes dissimilarities directly from the original variables. Because unsupervised dimension reduction has no ground-truth low-dimensional target, we assess performance through stability under perturbation. Across multiple datasets and analysis settings, the LLM embedding-based approach is consistently less stable than the direct tabular approach. In particular, small amounts of additional missingness and random permutation of feature order can substantially alter the resulting low-dimensional representation. These results suggest that the straightforward use of LLM text embeddings is not reliable for unsupervised dimension reduction of tabular data.

PMID:42577418 | PMC:PMC13454057 | DOI:10.3389/fbinf.2026.1851023

Ecol Evol. 2026 Aug 9;16(8):e74070. doi: 10.1002/ece3.74070. eCollection 2026 Aug.

ABSTRACT

Cuticular hydrocarbons (CHCs) are central to insect waterproofing and chemical communication, yet their diversity and comparative distribution in termites have not been synthesized comprehensively. Here, we present a systematic review of termite CHCs with emphasis on worker profiles, caste-level comparisons, and descriptive ecological interpretation. The qualitative synthesis included 28 studies, whereas the filtered statistical dataset comprised 37 worker profiles representing 37 species. Across the reviewed literature, termite cuticles contained a recurrent but variable combination of n-alkanes, n-alkenes, alkadienes, alkatrienes, mono-, di-, and trimethylalkanes. Species-level mapping showed that saturated and methyl-branched hydrocarbons were broadly distributed across the sampled taxa, whereas unsaturated compounds were more unevenly represented across the dataset. Genus-level synthesis showed that methyl-alkane-dominated, olefin-dominated, and mixed worker profiles occur in different higher taxonomic groups within the sampled dataset. Quantitative comparison of selected worker profiles highlighted broad chemistry in Mastotermes darwiniensis, strong methyl-branched dominance in Coptotermes formosanus, dimethyl-rich profiles in Nasutitermes corniger and N. ephratae, and marked alkatriene dominance in Parvitermes wolcotti. Caste-resolved studies showed that workers, soldiers, nymphs, and reproductives usually share the same broad CHC classes, with caste differentiation arising mainly through quantitative shifts, although reproductives in some taxa also express caste-associated compounds. Exploratory ecological analyses identified associations between nesting life type and worker CHC composition: drywood/enclosed wood nesters showed fewer dimethylalkanes and more frequent alkatrienes than other life types; but these patterns could not be separated from broad family-level phylogenetic structure. Overall, the available evidence suggests that termite CHC diversity reflects variation in the relative representation of a recurring chemical toolkit, with exploratory associations between worker CHC composition and nesting life type. Because the current dataset is limited and phylogenetic signal was not formally tested, these interpretations should be considered exploratory.

PMID:42577414 | PMC:PMC13454219 | DOI:10.1002/ece3.74070

Front Neurosci. 2026 Jul 27;20:1780430. doi: 10.3389/fnins.2026.1780430. eCollection 2026.

ABSTRACT

Autism Spectrum Disorder (ASD) is associated with atypical large-scale brain network organization, yet how spatial-functional dependencies relate to clinical features and molecular reference maps remains incompletely understood. To quantify spatial functional heterogeneity (Sill) and coherence persistence (Range), we analyzed resting-state fMRI data from 162 ASD and 175 TD adolescents, all aged 12-18. Compared with TD, adolescents with ASD exhibited significantly increased Sill within higher-order association networks, including the left Language and right Posterior Multimodal networks, whereas no group differences in Range survived multiple-comparison correction. Within the ASD group, elevated Sill was selectively associated with greater social-affective symptom severity but not restricted and repetitive behaviors. To explore potential biological correlates, we integrated cortical gene expression reference data and identified transcriptomic patterns associated with regional Sill differences. These genes showed enrichment for synaptic signaling, mitochondrial processes, and glial-related functions, highlighting multiscale correspondence between spatial-functional organization and molecular reference maps. Together, these results demonstrate statistical associations among altered spatial-functional properties, clinical severity, and transcriptomic profiles related to synaptic signaling, mitochondrial processes, and glial-related functions in ASD, providing a complementary spatial perspective on large-scale functional organization.

PMID:42577408 | PMC:PMC13454100 | DOI:10.3389/fnins.2026.1780430

Int J Public Health. 2026 Jul 27;71:1609229. doi: 10.3389/ijph.2026.1609229. eCollection 2026.

ABSTRACT

OBJECTIVES: School-based health promotion is critical for enhancing health literacy and academic performance. In South Africa, preventable visual impairment among school children may be exacerbated by the absence of structured eyecare promotion interventions. The objective of the study was to implement and evaluate the effectiveness of eyecare health promotion in the Thabo Mofutsanyane district, Free State province.

METHODS: A school-based cross-sectional interventional study was conducted. Following a simple randomised assignment, 10 schools received the intervention while the remaining 10 received no intervention. An adopted and piloted questionnaire was administered at baseline and readministered 6 months after the intervention. The McNemar’s chi-square test was used for statistical analysis.

RESULTS: The baseline study included 199 participants, and following attrition, stood at 136 learners, parents, and teachers. The learner experimental arm demonstrated a statistically significant change in eyecare knowledge (p < 0.05). While no statistically significant changes were noted among teachers, parents and learners had statistically significant variables.

CONCLUSION: Despite limitations, this study demonstrated that targeted eye health promotion can improve eyecare knowledge.

PMID:42577397 | PMC:PMC13453956 | DOI:10.3389/ijph.2026.1609229

Front Microbiol. 2026 Jul 27;17:1869903. doi: 10.3389/fmicb.2026.1869903. eCollection 2026.

ABSTRACT

INTRODUCTION: Conventional microbiological methods typically examine isolated bacterial species, which limits the insight into how microorganisms behave in more realistic, complex ecosystems. In contrast, synthetic bacterial consortia offer a practical and ecologically meaningful model for studying species interactions and responses to stress. These interspecies dynamics, both cooperative and competitive, can significantly influence the survival and physiology of individual species, revealing patterns that single-species tests often miss. Exploring microbial resilience under stress is vital for advancing space microbiology and ensuring the functionality of enclosed environments like the International Space Station (ISS) and in the future in spacecraft traveling to the Moon or Mars.

METHODS: To assess the stress tolerance of bacterial species within a consortium, a synthetic bacterial consortium was constructed using species representative of the spacecraft microbiome, including genera such as Bacillus, Pseudomonas, and Staphylococcus. The stress tolerance of Bacillus subtilis and Staphylococcus capitis was evaluated both individually and within the defined consortium under spaceflight-relevant conditions, including desiccation, X-ray irradiation, and hydrogen peroxide exposure. Survival rates were quantified by colony-forming unit (CFU) counts. Co-cultivation approaches were employed to assess biofilm formation and growth within the consortium. Consortium dynamics following stress exposure and cultivation were further investigated using 16S rRNA gene sequencing.

RESULTS: The stress tolerance of B. subtilis and S. capitis differed between individual exposure and consortium conditions. Hereby, the consortium context influenced stress responses in a stressor-specific manner. Co-cultivation experiments demonstrated observable growth in various combinations of consortium members, with differences in early-stage biofilm formation and CFU counts noted. Investigations of consortium dynamics revealed a statistically significant difference before and after incubation indicating that over time the consortium reaches a stabilized relative abundance profile.

DISCUSSION: Our findings underscore the importance of considering community-level interactions when evaluating bacterial stress tolerance. Synthetic bacterial consortia represent a valuable approach to bridging the gap between reductionist and systems-level microbiology, offering critical insights for both terrestrial biotechnology and the advancement of space exploration.

PMID:42577390 | PMC:PMC13454065 | DOI:10.3389/fmicb.2026.1869903

Front Bioeng Biotechnol. 2026 Jul 27;14:1787510. doi: 10.3389/fbioe.2026.1787510. eCollection 2026.

ABSTRACT

INTRODUCTION: Postural transitions (PTs) are crucial daily movements often impaired in neurological conditions, impacting autonomy and fall risk. Wearable inertial measurement units (IMUs) enable objective assessment of PTs, but robust algorithms for automatic identification remain limited.

METHODS: This pilot study used Dynamic Time Warping (DTW), a time-series alignment method that is robust to temporal variations, to automatically identify PTs in healthy subjects (HS) and subjects with Parkinson’s Disease (SwPD). For this purpose, 10 participants (5 HS and 5 SwPD) performed 5 postural transition tasks (sit-to-stand, stand-to-sit, supine-to-sit, sit-to-supine, and roll) using a sternum-mounted IMU. Reference PT patterns were generated from previously collected acceleration data representing optimally executed transitions. The DTW algorithm classified each detected candidate transition by minimizing its distance from predefined reference patterns. Classification performance indexes were statistically compared between groups.

RESULTS: Within this pilot dataset, the DTW algorithm correctly identified 118/118 postural-transition signals in HS and 136/147 signals in SwPD. Misclassifications in SwPD primarily affected sit-to-stand (31%) and stand-to-sit (19%) transitions.

DISCUSSION: These preliminary findings support the feasibility of a DTW-based approach for postural-transition identification, although larger independent validation studies are required. Future work should prioritize real-time deployment and larger validation cohorts.

PMID:42577388 | PMC:PMC13454294 | DOI:10.3389/fbioe.2026.1787510

Front Public Health. 2026 Jul 27;14:1827949. doi: 10.3389/fpubh.2026.1827949. eCollection 2026.

ABSTRACT

OBJECTIVE: To investigate the prevalence of gastrointestinal symptoms (reflux, nausea, and vomiting) among nurses in COVID-19 isolation wards and identify associated factors, with a specific focus on discomfort caused by personal protective equipment (PPE).

STUDY DESIGN: A cross-sectional survey.

METHODS: In March 2020, 354 of 368 eligible nurses (96.2% response rate) from the COVID-19 isolation wards of a designated hospital in Wenzhou, China, completed the survey. Data were collected using a demographic questionnaire, a self-rated PPE discomfort scale, the Gastroesophageal Reflux Disease Questionnaire (GerdQ), the Pittsburgh Sleep Quality Index (PSQI), and the Symptom Checklist-90 (SCL-90). Statistical assessments included univariate, correlation, and multivariable logistic regression analyses. Firth’s penalized likelihood logistic regression and sensitivity analyses were applied to address sparse data in specific subgroups.

RESULTS: The overall prevalence of gastrointestinal symptoms in the cohort was 23.2% (82/354). Multivariable regression identified several independent factors significantly associated with these symptoms (all P < 0.01): severe PPE-induced discomfort (OR = 3.64, 95% CI: 1.95-6.80), elevated psychological stress (SCL-90 total score: OR = 1.14, 95% CI: 1.09-1.20), and poor sleep quality (PSQI total score: OR = 2.10, 95% CI: 1.51-2.92). Although univariate analyses suggested protective associations for prior intensive care unit (ICU) experience and male gender, these effects lost statistical significance following Firth correction. Temporally, 76.8% of symptomatic cases emerged within the first five days of the shift cycle.

CONCLUSIONS: Gastrointestinal symptoms are prevalent among nurses in COVID-19 isolation wards and are strongly associated with PPE-induced discomfort, psychological stress, and sleep disturbances. Although the cross-sectional design precludes causal inference, these findings underscore an urgent need for the ergonomic optimization of PPE to alleviate heat stress and physical burden. Furthermore, implementing rapid adaptation training programs and integrating psychological and sleep support into routine occupational health surveillance are vital to safeguarding frontline clinicians.

PMID:42577379 | PMC:PMC13454333 | DOI:10.3389/fpubh.2026.1827949