Nevin Manimala

Stress Health. 2026 Feb;42(1):e70146. doi: 10.1002/smi.70146.

ABSTRACT

The COVID-19 pandemic has profoundly affected healthcare workers, increasing vulnerability to neuropsychiatric disorders, such as anxiety and depression. Psychological distress may be shaped by resilience, coping behaviours, and immune dysregulation. We investigated psychological distress symptoms, resilience, alcohol use, and cytokine profiles in 1440 workers from four hospitals in Fortaleza, Brazil. Participants were classified as frontline or second-line workers and assessed with the SRQ-20, CD-RISC, and AUDIT. Blood samples were analysed for SARS-CoV-2 antibodies and cytokines. Data were collected at two time points (August-October 2021; March-April 2022). Frontline workers reported higher distress, with decreased vital energy and somatic symptoms most prominent. Lower resilience scores correlated with all SRQ-20 domains, while higher alcohol use was linked to decreased energy and depressive thoughts. Reduced anti-spike antibody levels were also associated with greater distress. COVID-19 infection and symptom severity were associated with more persistent mental distress symptoms. Sex-specific immune signatures emerged: in women, lower interleukin (IL)-7 and C-X-C motif chemokine ligand 9 (CXCL-9) and higher IL-27 correlated with depressive-anxious mood and energy depletion; in men, IL-18, IL-9, and tumour necrosis factor beta (TNF-β) were positively associated with distress. This study demonstrates that psychological distress among healthcare workers during COVID-19 was shaped by resilience, alcohol use, infection severity, and sex-dependent immune alterations. Strengthening resilience and targeting inflammatory pathways may help mitigate the long-term mental health burden in this workforce during future public health crises.

PMID:41667934 | DOI:10.1002/smi.70146

Zhonghua Wai Ke Za Zhi. 2026 Feb 1;64(2):182-190. doi: 10.3760/cma.j.cn112139-20250814-00402.

ABSTRACT

Objective: To explore the performance of large language model (LLM) in diagnosing clinically significant prostate cancer (csPCa), and the improvement in diagnostic performance of open-source LLM after low-rank adaptation (LoRA) fine-tuning. Methods: This is a retrospective case series study. Data from 1 077 patients who underwent ultrasound-guided systematic prostate biopsy at Department of Urology,Peking University Third Hospital from January 2018 to December 2024 were collected, aged (M(IQR)) 69(13) years (range:38 to 90 years) including 391 patients in the gray zone (prostate-specific antigen 4 to 10 μg/L). The collected data included patients’ clinical characteristics, prostate MRI reports, and biopsy histopathological results. Four LLM (GPT 4.1, DeepSeek R1, Qwen3-235B-A22B, Qwen3-32B) were used to diagnose csPCa based on patient information, and the performance of the LLM was evaluated using biopsy histopathological results as the gold standard. Subsequently, the data from 1 077 patients were divided into training and test sets at an 8∶2 ratio, and LoRA fine-tuning was performed on Qwen3-32B. The fine-tuned model was named PCD-Qwen3, and its diagnostic efficacy in the test set was evaluated. The receiver operating characteristics curve was plotted and the area under the curve (AUC) and 95%CI were calculated to evaluate the diagnostic performance of LLM. The Delong test was used to compare the differences in AUC between groups. Results: Among all patients, DeepSeek R1 had the highest AUC for diagnosing csPCa at 0.848 (95%CI: 0.826 to 0.871), with statistically significant differences compared to Qwen3-235B-A22B (0.827 (95%CI: 0.803 to 0.851)) and Qwen3-32B (0.753 (95%CI: 0.724 to 0.781))(Z=2.34, P=0.020; Z=7.35, P<0.01), but no difference compared to GPT 4.1(0.842 (95%CI: 0.819 to 0.865))(P>0.05). The accuracy, sensitivity, and specificity of DeepSeek R1 for diagnosing csPCa were 77.3%, 70.2%, and 84.1%, respectively. In the gray zone patient population with total prostate specific antigen of 4 to 10 μg/L, DeepSeek R1 had an AUC of 0.765 (95%CI: 0.715 to 0.816) for diagnosing csPCa. Using DeepSeek R1 to diagnose gray zone patients could avoid 46.3% (181/391) of unnecessary biopsies while missing 5.9% (23/391) of csPCa patients. Except for Qwen3-32B, the PI-RADS scores evaluated by the three LLM achieved moderate agreement with those of radiologists. After LoRA fine-tuning, the diagnostic performance of PCD-Qwen3 was significantly improved compared to Qwen3-32B. In the test set of 216 patients, the accuracy, sensitivity, specificity, and AUC were 77.3%, 75.5%, 79.1%, and 0.831 (95%CI: 0.776 to 0.885), respectively, comparable to the performance of DeepSeek R1 (all P>0.05). Conclusions: Among the four LLM, DeepSeek R1 had the best performance in diagnosing csPCa. After LoRA fine-tuning, PCD-Qwen3 achieved performance comparable to DeepSeek R1. LLM demonstrated promising application value in diagnosing csPCa.

PMID:41667933 | DOI:10.3760/cma.j.cn112139-20250814-00402

Zhonghua Wai Ke Za Zhi. 2026 Feb 1;64(2):176-181. doi: 10.3760/cma.j.cn112139-20250920-00450.

ABSTRACT

Objective: To explore the perioperative blood loss characteristics of total knee arthroplasty (TKA) in high-altitude residents and identify the relative factors of blood loss. Methods: A retrospective cohort study was conducted, analyzing 400 patients with varus knee osteoarthritis who underwent unilateral TKA from January 2022 to May 2024 at the Second Affiliated Hospital of Army Medical University (Chongqing) and the 953 Hospital of People’s Liberation Army (Shigatse). There were 117 male cases and 283 female cases, with an age of (65.9±7.0) years (range:50 to 80 years). Based on the patients’ altitude of residence, they were divided into a high-altitude group (altitude of 3 800 meters, 200 cases) and a plain group (altitude of 360 meters, 200 cases). The perioperative bleeding conditions (total blood loss, intraoperative overt blood loss, postoperative drainage volume, and hidden blood loss) of the two groups of patients were compared. Through independent sample t test, χ² test, Pearson correlation analysis and multiple linear regression model, the influencing factors of perioperative blood loss in the plateau group were screened. Results: The intraoperative dominant blood loss of the plateau group was (219.7±108.4) ml, and the postoperative drainage volume was (378.8±144.8) ml. The corresponding values for the plainland group were (150.6±82.3) ml and (171.7±94.7) ml, respectively. The differences between the two groups were statistically significant(t=7.17, P=0.002; t=16.93, P<0.01). The GROSS equation calculated that the total perioperative blood loss of the plateau group was (1 144.9±367.4) ml, and the hidden blood loss was (545.5±299.2) ml, which were significantly higher than those of the plainland group (total blood loss (713.7±257.6) ml, hidden blood loss (387.6±257.4) ml), and the differences were statistically significant (t=13.59, P<0.01; t=5.66, P<0.01). Multivariate linear regression analysis revealed that gender, preoperative activated partial thromboplastin time (APTT), C-reactive protein (CRP), albumin, and bone density T value (all P<0.01) were independent influencing factors for perioperative blood loss in TKA in the plateau group. Conclusions: Perioperative blood loss in TKA for high-altitude residents is significantly higher than in plain areas. Gender, preoperative APTT, CRP, ALB, and bone density T value serve as independent relative factors.

PMID:41667932 | DOI:10.3760/cma.j.cn112139-20250920-00450

Cancer Epidemiol. 2026 Feb 9;101:103014. doi: 10.1016/j.canep.2026.103014. Online ahead of print.

ABSTRACT

BACKGROUND: Causes of childhood cancer remain poorly understood. Using data from the case-control studies of the Childhood Cancer and Leukemia International Consortium (CLIC), we explored how maternal education as a key socioeconomic status (SES) indicator, varies across studies/countries and contributes to understanding of potential environmental and lifestyle risk factors.

METHODS: Control group data from cancer-free children matched by diagnosis date of cases from 16 studies were included, using both interview-based and health registry sources. Maternal education, the primary SES measure used in previous analyses with pooled CLIC data, was categorized as low, medium, or high according to the International Standard Classification of Education. Multinomial logistic regression assessed associations between maternal education and perinatal/lifestyle factors, calculating crude and adjusted odds ratios (ORs) and 95 % confidence intervals (CIs) for high vs. low education.

RESULTS: Maternal education levels varied across studies and over time, with the highest proportions of highly educated mothers in the U.S. and lowest in Costa Rica, Italy, and Egypt. Higher maternal education was generally positively associated with higher birthweight, breastfeeding, daycare attendance, and maternal prenatal alcohol consumption. Higher maternal education was generally inversely associated with lower birthweight, younger maternal age, paternal occupational pesticide exposure, maternal prenatal smoking, and having more siblings. The direction of associations for older maternal age and for caesarean delivery differed substantially across regions. Exclusion of mothers < 21 years at birth of the index child had little effect on the results.

CONCLUSION: This multi-country analysis supports the use of maternal education for adjustment as a proxy for SES, showing largely consistent associations with various behaviors and exposures. While the direction of associations was generally consistent, the strengths varied sometimes considerably by geographical region. These findings support the inclusion of maternal education as a covariate in analyses of childhood cancer risk when pooling CLIC studies.

PMID:41666504 | DOI:10.1016/j.canep.2026.103014

Animal. 2026 Jan 20;20(3):101763. doi: 10.1016/j.animal.2026.101763. Online ahead of print.

ABSTRACT

Optimisation of slaughter weight is crucial for efficient farm management in all-in-all-out systems, but growth variability within pig batches complicates uniform marketing. This study aimed to reduce heterogeneity by developing a decision support system (DSS) for precision feeding, improving BW performance, thereby reducing batch variability. A 103-day commercial trial involving 365 pigs compared conventional 3-phase feeding with individual precision feeding (IPF). Two control groups, Control A and Control B (n = 81 each, six pens/group), received diets with stepwise standardised ileal digestible lysine (SID Lys) concentrations (8.80, 9.80, 10.60 g/kg and 9.00, 10.00, 10.80 g/kg, respectively) from traditional feeders, with feed intake recorded manually. In contrast, the IPF group (n = 203, nine pens) utilised robotic feeders to provide individually tailored diets. These were formulated in real-time by blending high (11.83 g/kg) and low (6.59 g/kg) SID Lys feeds. A DSS, integrating a nutritional model, stakeholder directives (minimum and limited daily decreases in the SID Lys concentration), and a qualitative model, calculated each pig’s requirements based on automatically collected real-time BW and feed intake data. Performance metrics were similar across all groups. However, the IPF group (18.55 g/kg) was more efficient in utilising SID Lys, requiring less per kg of live weight gain than Control A (19.67 g/kg) and Control B (19.71 g/kg). When pigs were classified by initial BW – heavy (HBW, IPF: 26; Control A: 23; Control B: 20 animals), moderate (MBW, IPF: 98; Control A: 39; Control B: 41 animals) and light-body-weight (LBW, IPF: 79; Control A: 19; Control B: 20 animals) – the IPF group showed an improvement of 4.2-6.8 kg in growth performance for HBW, and 2.6-4.3 kg in LBW, compared to controls, although not statistically significant. While overall batch variability remained similar (CV: 11.6% IPF, 11.9% Control A, 12.2% Control B), the IPF group was more homogeneous among LBW pigs (9.5%) compared to controls (11.5% and 13.8%). Greater HBW variation in IPF group balanced overall variability. Although direct feed cost savings and nitrogen excretion reductions were not achieved – attributed to technical feed distribution issues in the final phase and higher CP baselines in the experimental diets – an economic estimation revealed that the system’s profitability was driven by output maximisation. In conclusion, the DSS proved feasible for real-time commercial application, successfully enhancing nutrient utilisation efficiency and optimising the growth of animals at the extremes of the population distribution.

PMID:41666499 | DOI:10.1016/j.animal.2026.101763

Neural Netw. 2026 Feb 2;199:108676. doi: 10.1016/j.neunet.2026.108676. Online ahead of print.

ABSTRACT

Emotion recognition brain-computer interface (BCI) using electroencephalography (EEG) is crucial for human-computer interaction, medicine, and neuroscience. However, the scarcity of labeled EEG data limits progress in this field. To address this, self-supervised learning has gained attention as a promising approach. Despite its potential, self-supervised methods face two key challenges: (1) ensuring emotion-related information is effectively preserved, as its loss can degrade emotion recognition performance, and (2) overcoming inter-subject variability in EEG signals, which hinders generalization across subjects. To tackle these issues, we propose a novel knowledge-driven self-supervised learning framework for EEG emotion recognition. Our method incorporates domain knowledge to approximate the extraction of statistical feature differential entropy (DE), aiming to preserve emotion-related and generalizable information. The framework consists of two cascaded components as hard and soft alignments: a multi-branch convolutional differential entropy learning (MCDEL) module that simulates the DE extraction process, and a contrastive entropy alignment (CEA) module that exposes complex emotional semantics in high-dimensional space. Experiment results show that our method exhibits superior performance over existing self-supervised methods. The subject-independent mean accuracy and standard deviation of our method reached 84.48% ± 5.79 on SEED and 67.64% ± 6.35 and 68.63% ± 7.77 on the Arousal and Valence dimensions of DREAMER, respectively. We conduct an ablation study to demonstrate the contribution of each proposed component. Moreover, the t-SNE visualization intuitively presents the effect of our method on reducing inter-subject variability and discriminating emotional states.

PMID:41666485 | DOI:10.1016/j.neunet.2026.108676

Am J Trop Med Hyg. 2025 Dec 16:tpmd250128. doi: 10.4269/ajtmh.25-0128. Online ahead of print.

ABSTRACT

The epidemiological profile of hepatitis B (HBV) and hepatitis C (HCV) infections has not yet been sufficiently documented in the Democratic Republic of the Congo (DRC). The aim for the present study was to provide a descriptive analysis of HBV and HCV seroprevalence and assess trends, as well as any possible cohort effects in Bukavu, situated in the eastern DRC. Using laboratory data from the Provincial General Reference Hospital of Bukavu, the results of all HBV (hepatitis B surface antigen) and HCV (anti-HCV antibodies) serological tests performed between January 2019 and December 2023 were analyzed. Patients were grouped by possible complications and divided into age groups to assess trends and the cohort age effect. Of the 38,033 specimens tested, 807/19,333 (4.2%) and 321/18,700 (1.7%) tested positive for hepatitis B surface antigen and anti-HCV antibodies, respectively. Both infections were more prevalent in male participants than in female participants. The average age of patients was higher for those with HCV than for those with HBV (P = 0.0001). For HCV in particular, prevalence comparison between patients born before and after 1960 revealed a statistically significant difference: 10.5% versus 0.8% (P <0.0001). The profile analysis of the viral HBV and HCV epidemiology in Bukavu revealed significant changes over the years related to the degree of exposure to risk factors. These changes could explain the observed sex-related disparities regarding prevalence, as well as the cohort age effect clearly observed for HCV infections.

PMID:41666461 | DOI:10.4269/ajtmh.25-0128

Am J Trop Med Hyg. 2026 Feb 10:tpmd250329. doi: 10.4269/ajtmh.25-0329. Online ahead of print.

ABSTRACT

The Mandoto District in the central highlands of Madagascar experiences year-round transmission of Plasmodium vivax (P. vivax) and Plasmodium falciparum (P. falciparum). Monthly malaria case data from 27 health centers across Mandoto between 2019 and 2024 were analyzed alongside meteorological data to understand transmission dynamics and forecast potential influences of climate change using descriptive, cross-correlation, and seasonal autoregressive integrated moving average forecast models. Over a period of 6 years, 276,318 rapid diagnostic tests (RDTs) were performed, yielding a 39.6% positivity rate, totaling 109,428 malaria cases. After 2021, when multispecies RDTs became available, 71.5% of cases were attributed to P. falciparum, and 28.5% were attributed to P. vivax. Both species were co-endemic across all health centers, with the western region experiencing a higher transmission risk. Malaria cases peaked in January, with a second peak from April to June after the rainy season, and declined between July and September. Precipitation and temperature effectively revealed the seasonality of malaria dynamics, thereby improving model accuracy. Plasmodium falciparum exhibited stronger associations with precipitation and temperature variability. The present study highlights that combining time-series modeling with precipitation and temperature data can help predict malaria cases and support timely planning and resource allocation.

PMID:41666420 | DOI:10.4269/ajtmh.25-0329

Am J Trop Med Hyg. 2026 Jan 15:tpmd250372. doi: 10.4269/ajtmh.25-0372. Online ahead of print.

ABSTRACT

Soil-transmitted helminths (STHs) predominantly affect resource-poor communities with poor access to water, sanitation, and hygiene (WASH) infrastructure. Understanding the risk factors for STH infections can help inform public health control strategies, including WASH interventions and preventive chemotherapy. In this school-based cross-sectional study, mixed-effects logistic regression was used to examine the associations between WASH and socioeconomic factors and STH infections in Timor-Leste. Two statistical analyses were conducted: the first included individual-level sanitation and hygiene factors, whereas the second also included household-level WASH and socioeconomic factors. In the sanitation and hygiene analysis, “always use household latrine” was associated with lower odds of undifferentiated STH infection (adjusted odds ratio [aOR]: 0.59; 95% CI: 0.37-0.96). “Always wash hands before eating” was associated with lower odds of Trichuris trichiura (T. trichiura) infection (aOR: 0.35; 95% CI: 0.13-0.97), whereas “always have soap to wash hands at home” was associated with higher odds of T. trichiura infection (aOR: 4.22; 95% CI: 1.56-11.43). In the WASH and socioeconomic factors analysis, “usually defecate at household/neighbor’s latrine” was associated with lower odds of undifferentiated STH (aOR: 0.13; 95% CI: 0.04-0.43) and Ascaris lumbricoides infections (aOR: 0.19; 95% CI: 0.06-0.64). Additionally, the availability of school handwashing stations was associated with lower odds of T. trichiura infection (aOR: 0.21; 95% CI: 0.05-0.86). The present study indicates that sanitation and hygiene are important risk factors for STH infections, and therefore, efforts to reduce STH infections should also promote sanitation and hygiene infrastructure and practices.

PMID:41666419 | DOI:10.4269/ajtmh.25-0372

Brief Bioinform. 2026 Jan 7;27(1):bbag053. doi: 10.1093/bib/bbag053.

ABSTRACT

To understand how the tumor immune microenvironment (TIME) impacts clinical outcomes and treatment response, researchers have been leveraging single-cell protein multiplex imaging techniques. These technologies measure multiple protein markers simultaneously within a tissue sample, providing a more complete assessment of the TIME. However, statistical challenges arise from the over-dispersed and zero-inflated nature of the data and from relationships among different immune cell populations. To address these challenges, we developed a Bayesian hierarchical method using a beta-binomial (BB) distribution to model the abundance of multiple immune cell types simultaneously while incorporating relationships and immune cell differentiation paths. We applied the model to data from three large studies of high-grade serous ovarian tumors (Nurses’ Health Study I/II: N = 321, African American Cancer Epidemiology Study: N = 92, University of Colorado Ovarian Cancer Study: N = 103). We examined associations between cancer stage, age at diagnosis, and debulking status and the abundance of immune cell populations. We compared the multi-cell type model to individual cell type analyses using a Bayesian BB model. The multi-cell type model detected more associations, when present, with narrower credible intervals. To support broader application, we developed an R package, BTIME, with a detailed tutorial. In conclusion, the Bayesian multi-cell type model is flexible in how relationships between cell types are incorporated and can be used for cancer studies that interrogate the TIME.

PMID:41666406 | DOI:10.1093/bib/bbag053