Category: Statistics

PLoS Genet. 2026 Mar 24;22(3):e1012067. doi: 10.1371/journal.pgen.1012067. eCollection 2026 Mar.

ABSTRACT

Interbreeding between anatomically modern humans and archaic hominins has contributed to the genomes of present-day human populations. However, our understanding of the specific gene regulatory consequences of Neanderthal, and particularly, Denisovan introgression is limited. Here, we used a massively parallel reporter assay to investigate the regulatory effects of 25,869 high-confidence introgressed SNPs segregating in present-day individuals of Papuan genetic ancestry in immune cell types. Overall, 8.22% of Denisovan and 8.58% of Neanderthal sequences showed active regulatory activity, and 9.22% of these displayed differential activity between archaic and modern alleles. We found no association between introgressed allele frequency on activity regardless of introgression source, but introgressed Denisovan alleles at higher frequencies were less likely to be differentially active than expected, suggesting introgression is under some degree of selective constraint. Both activity and differentially activity were associated with distance to the nearest transcription start site, while differential activity was additionally associated with differential transcription factor binding. Genes predicted to be regulated by differentially active sequences included IFIH1 and TNFAIP3, key immune genes and known examples of archaic introgression. Overall, this work provides experimental validation of regulatory activity for thousands of archaic variants in populations with the highest levels of Denisovan ancestry worldwide, revealing how human evolutionary history actively shapes present-day genetic diversity and immune function.

PMID:41875405 | DOI:10.1371/journal.pgen.1012067

JMIR AI. 2026 Mar 24;5:e70017. doi: 10.2196/70017.

ABSTRACT

BACKGROUND: Despite the growing potential of large language models (LLMs) in mental health services, evidence on its capabilities in diagnostic processes remains limited.

OBJECTIVE: This study described the development and evaluation of CapyEngine, an LLM-powered diagnostic tool designed to assist in the diagnosis of mental disorders.

METHODS: We developed and evaluated CapyEngine through 3 phases. In phase 1, we created a disorder and symptom database using Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, Text Revision (DSM-5-TR). We then designed and developed CapyEngine’s architecture using LLMs, embedding models, and vector searches. In phase 2, we conducted usability testing with mental health professionals (n=7). In phase 3, we compared CapyEngine’s diagnostic accuracy against ChatGPT-4o and clinicians using 35 standardized case scenario exam questions from psychiatry and clinical psychology board exams. Questions were input into CapyEngine, and the top 10 recommended diagnoses were obtained. ChatGPT-4o was prompted to provide the top 10 potential diagnoses for each question. Clinicians (n=3) received similar instructions to generate at least 10 potential diagnoses for each question. Responses were then analyzed to compare diagnostic accuracy of CapyEngine, ChatGPT-4o, and clinicians. Accuracy was measured by the percentage of questions where the correct answer was among the top 10 (least stringent), top 5, or top 1 (most stringent) results of the diagnosis list.

RESULTS: Preliminary user interview reflected high acceptability and feasibility of CapyEngine. Across diagnostic accuracy thresholds, ChatGPT-4o consistently outperformed both CapyEngine and clinicians in broader rankings (top 10 and top 5 benchmarks; all P<.03). Clinicians showed significantly higher accuracy than CapyEngine using the top 5 benchmark (odds ratio 0.26, 95% CI 0.09-0.78; P=.02). For the top 1 benchmark, no significant differences were observed, where clinicians showed a borderline advantage over ChatGPT-4o (odds ratio 0.34, 95% CI 0.13-0.91; P=.05). Regarding the range and slope of diagnostic accuracy decline across benchmarks (least to most stringent), CapyEngine showed the smallest decline (0.14) and flattest slope (-0.07), reflecting more consistent and constrained diagnostic ranking behavior as evaluation thresholds became more stringent. Clinicians exhibited a moderate decline (0.26), whereas ChatGPT-4o demonstrated a sharp decrease (0.69) in accuracy when only the top-ranked diagnosis was considered, consistent with broader diagnostic coverage at less stringent thresholds.

CONCLUSIONS: Overall, ChatGPT-4o achieved the highest accuracy at less stringent benchmarks (top 10 and top 5), while clinician performance did not differ significantly from ChatGPT-4o in identifying the single most likely diagnosis. Although CapyEngine was less accurate overall, it exhibited more consistent and constrained diagnostic ranking across evaluation benchmarks, likely reflecting its DSM-5-TR-based, domain-specific design rather than broader diagnostic coverage. Nonetheless, CapyEngine shows promise as a tool to augment the mental health diagnostic process, and further research is needed to evaluate the risks and benefits of integrating artificial intelligence systems, such as CapyEngine, into clinical workflows.

PMID:41875403 | DOI:10.2196/70017

PLoS One. 2026 Mar 24;21(3):e0344023. doi: 10.1371/journal.pone.0344023. eCollection 2026.

ABSTRACT

Aiming at the TTC (Time to Collision) and derivative indicators’ problems of unclear definition and missed/wrong judgments of traffic conflicts on curved road in the traditional Cartesian coordinate system, a new method that can better identify the conflicts on curved road is proposed. The method first establishes the Frenet coordinate system according to the road centerline (i.e., the reference line), and obtains the vehicle trajectory coordinates in the Frenet coordinate system. The Frenet coordinate system can simplify the calculation difficulty of vehicle trajectory and conflict under the curve road. Then determine the vehicle state in the Frenet coordinate system, and then use TTC to calculate rear-end and lane-change conflicts according to the state of the vehicle (non-lane-change/lane-change). Finally, a total of 4 hours of video data were collected based on the K283 of the lane-switch work zone of the Jiqing Highway. Subsequently, the continuous high-precision conflict data in the region was obtained through the video and conflict identification program, and the traditional method was compared with the new method. The results show that different methods have a significant impact on the identification of the number of serious conflicts. The new method can reduce the missed judgments of serious rear-end conflicts on curved road, especially at the junctions of curved and straight segments (segment 3/4/7/8/9), and can also reduce wrong judgments of serious lane-change conflicts. In addition, among the 125 added serious rear-end conflicts identified by the new method, the maximum deceleration of 10 conflicting vehicles during the conflict exceeds the dangerous state -4/-1.5m/s2, which explain that the new method can help us better identify the risks of curved road. The new method combines the Frenet coordinate system, vehicle state determination and TTC, which can reduce the missed/wrong judgments of conflicts on curved road, and expand the traffic conflict identification from previous straight road to full-line road alignment.

PMID:41875402 | DOI:10.1371/journal.pone.0344023

PLoS One. 2026 Mar 24;21(3):e0344150. doi: 10.1371/journal.pone.0344150. eCollection 2026.

ABSTRACT

Acute meningitis remains a primary global health concern. Immunosuppressed patients have risks due to atypical clinical presentations and a broader range of causative pathogens. We aimed to describe the outcomes and clinical features of acute meningitis according to immunosuppression status. We performed a retrospective cohort study of adults with acute meningitis from January 2009 to December 2023. Patients with postsurgical meningitis were excluded. Outcomes and demographic, clinical, and laboratory features were compared using non-parametric statistical tests, and mortality was analyzed using multivariate logistic regression. Among 189 patients, 96 (51%) were immunosuppressed. The median age was lower in immunosuppressed patients (36 vs. 50 years, p < 0.01). There were no differences in symptoms; the classical triad was present in only 21% vs. 19%. Immunosuppressed patients had lower CSF glucose levels (59% vs. 39%, p = 0.004). Overall mortality was 20%, with no significant difference by immune status. Independent predictors of death included age over 50 years (OR 2.9), altered mental status (OR 4.7), and bacterial meningitis (OR 2.3). Acute meningitis in immunosuppressed hosts shows attenuated inflammatory CSF profiles and a broader etiologic spectrum. Immunosuppression was not independently associated with in-hospital mortality.

PMID:41875396 | DOI:10.1371/journal.pone.0344150

Interdiscip Cardiovasc Thorac Surg. 2026 Mar 24:ivag088. doi: 10.1093/icvts/ivag088. Online ahead of print.

ABSTRACT

OBJECTIVES: Abdominal aortic aneurysm (AAA) is often fatal when ruptured and current guidelines suggest surgical management at suprathreshold sizes (50 mm for women or 55 mm for men) or with rapid expansion (>5 mm/year). Many medical therapies have been assessed for reducing subthreshold AAA expansion though the evidence remains inconclusive. This network meta-analysis (NMA) compares AAA growth and mortality amongst medical treatments for AAAs.

METHODS: MEDLINE (via PubMed), Scopus, Web of Science, EBSCO, and Cochrane Library databases were searched for relevant randomized controlled trials (RCTs) from database inception to 2024. Outcomes assessed included AAA growth rate, rate of referral for aneurysm surgery, overall mortality, and discontinuation from adverse effects. Data was analyzed using R software, and P-score was used to rank different treatments. The GRADE framework was performed to assess quality of evidence.

RESULTS: Thirteen RCTs comprising 3084 patients were included in this NMA. AAA diameters ranged from 3.1-4.6 cm in the intervention group and 3.5-4.5 cm in the placebo group. Study-level mean annual growth rate ranged from 1.2-2.8 mm/year in the intervention group compared with placebo (1.2-2.6 mm/year). There were no significant differences in AAA growth among the compared groups, (P-score probability in brackets): propranolol (0.73) telmisartan (0.66), antibiotics (0.53), placebo (0.53), ACE inhibitors (0.52), ticagrelor (0.46), and pemirolast (0.06). There were no significant differences among the compared groups in terms of aneurysm surgery referral rates, with propranolol (0.91), antibiotics (0.56), placebo (0.45), and pemirolast (0.08) showing similar outcomes. Similarly, no significant differences were observed in overall mortality rates across the groups, including telmisartan (0.87), antibiotics (0.57), ACE inhibitors (0.51), placebo (0.35), and propranolol (0.17). However, propranolol (OR = 3.14, 95% CI [1.34, 7.35]) and ticagrelor (OR = 5.10, 95% CI [1.12, 23.18]) were associated with a higher rate of discontinuation due to adverse events. Most of the studies analysed demonstrate moderate quality evidence.

CONCLUSIONS: Current evidence highlights ongoing uncertainty regarding the efficacy of medical therapies in reducing subthreshold AAA growth rates, rates of referral for surgical repair, or overall mortality. The absence of statistically significant benefit may reflect underpowered datasets rather than definitive treatment inefficacy. Future large-scale, appropriately powered randomized controlled trials evaluating emerging medical treatments are required to accurately assess their clinical potential.

PMID:41875387 | DOI:10.1093/icvts/ivag088

NCHS Data Brief. 2026 Mar;(552). doi: 10.15620/cdc/174648.

ABSTRACT

INTRODUCTION: This report provides the most recent prevalence estimates of abnormal blood cholesterol in children and adolescents ages 6-19 and describes changes in prevalences over time.

METHODS: Cholesterol, anthropometry, and demographics data from the National Health and Nutrition Examination Survey (NHANES) for 2013-2014 through August 2021-August 2023 were used for these analyses. Phlebotomy sample weights were used to estimate prevalence, and confidence intervals were estimated using Taylor series linearization. Statistically significant differences in prevalence estimates by age, sex, and weight status were tested using a t statistic at the p < 0.05 level, and trends were evaluated using linear regression models.

KEY FINDINGS: During August 2021-August 2023, 16.5% of children and adolescents had at least one measure of abnormal cholesterol (high total cholesterol, low high-density lipoprotein cholesterol [HDL-C], or high non-HDL-C). The prevalence of at least one abnormal cholesterol measure was lower in girls (13.6%) than in boys (19.2%) and lower in those with underweight or normal weight (10.3%) or overweight (11.5%) than in those with obesity (35.8%). The prevalence of at least one abnormal cholesterol measure decreased between 2013-2014 (21.3%) and August 2021-August 2023 (16.5%).

PMID:41875385 | DOI:10.15620/cdc/174648

Genetics. 2026 Mar 24:iyag081. doi: 10.1093/genetics/iyag081. Online ahead of print.

ABSTRACT

In many taxa, levels of genetic diversity vary along their genomes. The framework of background selection models this variation in terms of linkage to constrained sites, and recent applications have been able to explain a large portion of the variation in human genomes. However, these studies have also yielded conflicting results, stemming from two key limitations. First, existing models are inaccurate in a critical region of parameter space (Nes ∼ -1), where the local reduction in diversity is sharpest. Second, they assume a constant population size over time. Here, we develop predictions for diversity under background selection based on the Hill-Robertson system of two-locus statistics, which allows for population size changes. We treat the joint effect of multiple selected loci independently, but we show that interference among them is well captured through local rescaling of mutation, recombination and selection in an iterative procedure that converges quickly. We further accommodate existing background selection theory to non-equilibrium demography, bridging the gap between weak and strong selection. Simulations show that our predictions are accurate across the entire range of selection coefficients. We characterize the temporal dynamics of linked selection under population size changes and demonstrate that patterns of diversity can be misinterpreted by other models. Specifically, biases due to the incorrect assumption of equilibrium carry over to downstream inferences of the distribution of fitness effects and deleterious mutation rate. Jointly modeling demography and linked selection therefore improves our understanding of the genomic landscape of diversity, which will help refine inferences of linked selection in humans and other species.

PMID:41875378 | DOI:10.1093/genetics/iyag081

JMIR Form Res. 2026 Mar 24;10:e83430. doi: 10.2196/83430.

ABSTRACT

BACKGROUND: Compared with other mental health problems, self-directed interventions for gambling problems lack in quantity, accessibility, and in some cases, evidence base. Moreover, engagement with these interventions remains modest. Mobile apps may be a viable format to deliver self-directed interventions that enhance user engagement.

OBJECTIVE: The aims of this study were to develop a self-directed mobile app intervention for gambling problems and to conduct initial feasibility and acceptability testing with a small sample of Canadian adults with past or present gambling problems (n=30).

METHODS: Participants were invited via email from a list of people who had previously volunteered in similar research in our laboratory. Theory and content of the mobile app intervention were primarily based on a self-directed workbook that has been evaluated in paperback and static web-based formats. The current app prototype included daily gambling diaries, recommended activities based on diary responses, and psychoeducation. It was available for 2 weeks, after which users provided feedback via surveys (n=30) and a virtual focus group (n=8). Quantitative and qualitative feedback, as well as app usage data, were analyzed to provide descriptive statistics and summaries.

RESULTS: Regarding feasibility, median completion time for activities ranged from 48 (IQR 35-90) to 137 (IQR 93-328) seconds. Daily diary completion rate was 51%. One-third of activities were accessed via prompt, and two-thirds on demand. Many participants repeated at least 1 activity, and all activities were repeated by at least 1 participant. Results also indicated favorable user reviews, particularly regarding the app’s credibility, ease of use, and potential impact. The feedback on some app features was highly variable, such as the perceived utility of daily diaries. Specific recommendations for improvement were provided, such as the inclusion of information on concurrent substance use and more interactive psychoeducation.

CONCLUSIONS: Overall, the app met or exceeded heuristic thresholds for feasibility and acceptability testing. These results will inform improvements and subsequent effectiveness testing. The variability in user feedback underscores the demand for further personalization.

PMID:41875357 | DOI:10.2196/83430

J Public Health Manag Pract. 2026 May-Jun 01;32(3):414-419. doi: 10.1097/PHH.0000000000002285. Epub 2025 Nov 24.

ABSTRACT

The 2019-2020 and 2023-2024 flu seasons had unusually active influenza activity late in the season. As vaccines can be a cost-effective way of avoiding severe illness with the flu resulting in hospitalizations or mortality, the Louisiana Department of Health mailed reminder-recall postcards to adults aged 65 to 70 years to raise awareness about respiratory vaccines, including those for flu. In the 2 weeks following the distribution of reminder-recall mailers (n = 47 587 in February 2020 and n = 42 364 in January 2024), an additional 1017 and 1093 flu shots were administered in 2020 and 2024, respectively. These doses comprised 1% of all flu shots administered in their respective seasons and 2% of people who received the mailers. Together, the results suggest potentially limited value of flu reminder-recall mailers for this population without further research on timing, frequency, and cost-effectiveness of the same.

PMID:41875352 | DOI:10.1097/PHH.0000000000002285

Gastroenterol Nurs. 2026 Mar-Apr 01;49(2):74-85. doi: 10.1097/SGA.0000000000000948. Epub 2026 Mar 23.

ABSTRACT

Self-efficacy in coping with cancer surgery is a key concept for planning and organizing nursing care. However, there is no validated measurement tool that can be used to assess self-efficacy in patients with colorectal cancer surgery. This cross-sectional study, with a multi-phase design, was conducted with 170 patients to describe the development and validation of the Self-Efficacy Scale for Patients with Colorectal Cancer Surgery (SES-CRCS). Phase 1 addressed the development procedures of the scale. In Phase 2, a 3-step validation process was conducted: (a) assessing the content validity, (b) evaluating construct validity with exploratory and confirmatory factor analysis, and (c) assessing internal consistency reliability with Cronbach’s α coefficient, test-retest, and item-total correlation methods. Exploratory factor analysis suggested a 16-item single-factor structure. The factor loadings of the 16 items were above .40, and various indices used to examine the consistency of the scale indicated a good model fit. The Cronbach’s α value was .838, the item-total correlations for all items were positive, and there was a strong correlation between the test-retest measurements (r = .933; p < .05). As a result, the SES-CRCS was found to be a valid and reliable measurement tool in assessing self-efficacy among patients with colorectal cancer surgery.

PMID:41875333 | DOI:10.1097/SGA.0000000000000948