Tag: statistics

Acta Med Acad. 2025 Nov 20. doi: 10.5644/ama2006-124.488. Online ahead of print.

ABSTRACT

OBJECTIVE: This study analyzed PhD theses defended in Croatia between 1992 and 2023, with the aim of examining national trends, institutional contributions, disciplinary patterns, and data-related challenges.

METHODS: This retrospective time-trend study utilized the administrative data obtained from the Croatian Bureau of Statistics. Data on the number of defended PhD theses were collected by year, university, and school/department. Linear regression models were applied to assess temporal trends at both the national and institutional levels.

RESULTS: A total of 17,578 PhD theses were defended in Croatia between 1992 and 2023. The national output increased substantially, reaching a peak of 1,338 theses in 2012, followed by a subsequent decline and a gradual recovery. The University of Zagreb accounted for 74.8% of all defended theses, followed by the Universities of Osijek, Rijeka, and Split. Across institutions, the medical, economic, and engineering faculties were the most productive. Linear regression analyses demonstrated statistically significant upward trends at both the national level and across all major public universities. Collectively, medical schools produced 18% of all theses, with newer institutions, particularly those in Split and Osijek, exhibiting later but consistent growth. However, notable data inconsistencies were observed, including non-standardized institutional nomenclature, variable data granularity, and discrepancies among official reports.

CONCLUSION: Croatia’s PhD output expanded markedly after 2000, reflecting the maturation and expansion of its higher education system. Regional universities and medical schools substantially increased their contributions, indicating national academic growth. Sustained institutional support will be essential to sustain progress and foster disciplinary development.

PMID:41268749 | DOI:10.5644/ama2006-124.488

Per Med. 2025 Nov 21:1-9. doi: 10.1080/17410541.2025.2592528. Online ahead of print.

ABSTRACT

BACKGROUND: Urinary bladder cancer (UBC), the tenth most common globally with a male predominance, has its risk factors like smoking and single nucleotide polymorphisms (SNPs), yet the MYC gene’s role, especially the rs9642880 GT/TT polymorphism on chromosome 8q24.21, in cancer susceptibility is underexplored.

METHODS: We searched in Medline, Scopus and Web of Science databases for relevant studies. Odds ratios (OR) with 95% confidence intervals (CIs) were computed using DerSimonian and Laird random-effect models. Heterogeneity was assessed using I² statistics. Statistical analyses were performed using R, version 4.4.2.

RESULTS: Data from 53,957 subjects was analyzed, including 8,002 UBC patients and 45,955 controls, with median age ranging from 44.5 to 68.7 years. We found significant association between the rs9642880 GT/TT polymorphism and increased UBC risk (OR 1.249, 95% CI 1.188-1.297, p < 0.001). This association was consistent across American (OR 1.43 95% CI 1.20-1.70), Asian (OR 1.25 95% CI 1.17-1.34), and Caucasian (OR 1.21 95% CI 1.13-1.29) ethnic groups, with low heterogeneity (I² = 0%). Subgroup analyses by genotyping method and source of control further supported these results.

CONCLUSION: This meta-analysis provides robust evidence that the T allele of the rs9642880 polymorphism in the MYC gene significantly increases susceptibility to UBC across Asian, American, and Caucasian populations.

PMID:41268746 | DOI:10.1080/17410541.2025.2592528

J Med Virol. 2025 Nov;97(11):e70719. doi: 10.1002/jmv.70719.

ABSTRACT

Genome-wide association studies (GWAS) have identified numerous genetic loci associated with COVID-19 susceptibility and severity. However, it remains unclear which variants specifically contribute to the progression from hospitalization to critical illness. To address this, we identified 394 symptom-differentiating SNPs (sdSNPs) by directly comparing GWAS summary statistics between critical and hospitalized COVID-19 cases. Among these, 13 were missense variants that may trigger phenotype conversion. To investigate their functional consequences, we performed structural modeling to evaluate the impact of these variants on protein stability. Eight proteins were predicted to be affected, including FUT2, ICAM5, IFNA10, PLSCR1, IZUMO1, ICAM1, RASIP1, and MICB, suggesting potential disruptions in immune signaling and host response pathways. Furthermore, Mendelian randomization analyses provided direct causal evidence: SMR analysis based on lung eQTL data demonstrated that higher FUT2 expression was significantly associated with increased risk of critical COVID-19 (β = 0.227, p = 2.69 × 10^-3), compared to a weaker effect in hospitalized COVID-19 (β = 0.096, p = 3.97 × 10^-2). In contrast, MR analysis using cis-pQTLs revealed that higher plasma levels of soluble ICAM5 (sICAM5) exerted protective effects, with stronger estimates in critical COVID-19 (β = -0.186, p = 5.54 × 10^-7) than in hospitalized COVID-19 (β = -0.095, p = 1.69 × 10^-4). These findings identify a distinct set of severity-specific genetic variants and offer mechanistic insights into how missense mutations may influence disease progression through structural and regulatory pathways.

PMID:41268733 | DOI:10.1002/jmv.70719

Psychol Med. 2025 Nov 21;55:e355. doi: 10.1017/S0033291725102286.

ABSTRACT

BACKGROUND: Individuals with first-episode psychosis (FEP) face markedly increased excess mortality, yet the long-term trends and key contributing factors remain insufficiently characterized. This study aimed to examine long-term mortality patterns, standardized mortality ratios (SMRs), and associated factors in a FEP cohort.

METHODS: This population-based cohort study included 1,389 individuals diagnosed with FEP, followed for up to 25 years. Mortality outcomes were obtained from Hong Kong’s centralized hospital database (CMS) and coroner’s court reports, with SMRs calculated. Baseline sociodemographic and clinical, as well as long-term treatment-related factors of all-cause, natural, and unnatural mortality were analyzed.

RESULTS: Among 1,389 participants, 137 deaths (9.86%) occurred during the follow-up period with the overall SMR of 6.56 (95% CI, 5.50-7.71). The cumulative incidence rate of unnatural mortality increased sharply over the first 10 years and that of the natural cause of death started to increase after the first decade of the illness. Male gender and poorer social functioning were associated with increased all-cause mortality risk, while male gender, lower education, and baseline hospitalization raised unnatural mortality risk. Greater monthly antipsychotic variability during the first 10 years increased all-cause mortality risk in the period after the initial 10 years.

CONCLUSIONS: This 25-year follow-up study of FEP highlighted the changes in the long-term mortality pattern of FEP and thus the phase-specific needs of individuals with FEP. Therefore, it is important to integrate physical care into mental health services, as well as stage-specific and individualized care for patients with psychotic disorders to reduce long-term excess mortality.

PMID:41268721 | DOI:10.1017/S0033291725102286

Child Care Health Dev. 2025 Nov;51(6):e70187. doi: 10.1111/cch.70187.

ABSTRACT

BACKGROUND: Autistic students experience many problems with school attendance. School exclusion and truancy are among the least researched school attendance problems in this population. The study aimed to describe levels of exclusion and truancy in a UK population-representative sample of autistic adolescents and identify child, family and school factors associated with each school attendance problem.

METHODS: Data were drawn from the Millennium Cohort Study where exclusion and truancy information was available for 460 autistic 14-year-olds. Descriptive statistics were used to report the weighted prevalence of exclusion and truancy. A bio-ecological framework guided the selection of available child, family and school factors potentially associated with exclusion and truancy for modelling. Univariable and multivariable logistic regression models were fitted to investigate associations.

RESULTS: Twenty-five percent of autistic adolescents were temporarily or permanently excluded at least once. Fifteen percent of autistic adolescents reported truanting at least once. Externalising problems were strongly associated with higher odds of exclusion and truancy. The level of school support was strongly associated with higher odds of exclusion. There was weak evidence of an association between exclusion and low parental school engagement and between truancy and the absence of intellectual impairment.

CONCLUSION: Poor mental health and in particular externalising difficulties seem to be strongly associated with a greater likelihood of parent-reported exclusion and adolescent-reported truancy. The role of school support and adaptation to the child’s needs warrants further investigation.

PMID:41268717 | DOI:10.1111/cch.70187

Antivir Ther. 2025 Dec;30(6):13596535251392955. doi: 10.1177/13596535251392955. Epub 2025 Nov 21.

ABSTRACT

BackgroundClass E (empty) capsid assembly modulators (CAM-Es) inhibit HBV capsid assembly, pregenomic RNA encapsidation preventing formation the establishment of covalently closed circular HBV DNA (ccDNA). ALG-000184 (pevifoscorvir sodium), is a prodrug of the Class E CAM ALG-001075.MethodsALG-000184-201 was a Phase 1 randomized, double-blind, placebo-controlled study to evaluate the safety, tolerability and pharmacokinetics of ALG-000184. Healthy participants (n = 8/cohort) received oral single-ascending doses (SAD) of ALG-000184 (40, 100, 250, and 500 mg) or placebo, and multiple-ascending daily doses (MAD) (150 mg and 250 mg) or placebo for 7 days.ResultsALG-000184 was well tolerated by 48 participants who received single doses up to 500 mg and multiple daily doses up to 250 mg for 7 days. ALG-000184 was rapidly converted to the active moiety, ALG-001075. ALG-001075 had dose-proportional increase in plasma exposure, low-to-moderate variability (18%-34% CV for AUC_0-24), rapid absorption (median t_max 1-3.5 h), and biphasic distribution/elimination with terminal t½ 7-8 h and minimal accumulation (∼30%). A major oxidative metabolite, ALG-000302, was identified in plasma (∼17%-24% of ALG-001075). A high-fat/high-calorie meal did not significantly impact the plasma pharmacokinetics. No differences in pharmacokinetics between Asian and non-Asian participants were observed. A concentration QT analysis indicated no statistically significant change in ΔΔQTcF with plasma ALG-001075. Urinary excretion of ALG-001075 was low following single or multiple ALG-000184 doses.ConclusionsALG-000184 demonstrated good tolerability, safety and pharmacokinetic properties in healthy participants. The pharmacokinetic profile suggests that a daily dose of 100 mg or higher will provide efficacious exposures in patients with chronic HBV infection.Clinical trial numberNCT04536337 (https://clinicaltrials.gov/study/NCT04536337).

PMID:41268712 | DOI:10.1177/13596535251392955

Int J Phytoremediation. 2025 Nov 21:1-14. doi: 10.1080/15226514.2025.2585364. Online ahead of print.

ABSTRACT

Plant-colonizing beneficial microbes are effective bio-tools for enhancing phytoremediation. Two-year pot experiment at Punjab Agricultural University, Ludhiana, India, assessed the response of Populus deltoides under nursery conditions to sewage sludge treated soil with indigenous metal-mobilizing Bacillus species-B. thuringiensis (T₁), B. cereus (T₂), B. pumilus (T₃), and their consortium (T₄), with three inorganic fertilizer levels-RDF1-100%, RDF2-75%, and RDF3-50% [Recommended dose of fertilizer (RDF)]. Each inoculated treatment was compared to its respective uninoculated control (C). The application of T₄ with RDF1 significantly increased shoot length and biomass by 13.8 and 32.9% than C, respectively. Bioconcentration factors (BCF) for Cd and Ni increased by over 50% than C demonstrating enhanced phytoremediation efficiency. Elemental accumulation was predominantly localized in roots, with the exception of Zn and Cd. Among most of the parameters, RDF1 × T₄ was statistically comparable with RDF2 × T_4. Irrespective of fertilizer dose, T₄ maximally improved phytoremediation efficiency (BCF) by 0.61 (shoot) and 0.52 (root) compared to 0.20 and 0.16 in C, respectively, as well as soil chemical and biological properties up to 22.3%. These results highlight the potential of indigenous microbial inoculants to reduce soil heavy metals and enable sustainable, enhanced phytoremediation with 25% lower fertilizer input.

PMID:41268680 | DOI:10.1080/15226514.2025.2585364

J Child Adolesc Psychiatr Nurs. 2025 Nov;38(4):e70045. doi: 10.1111/jcap.70045.

ABSTRACT

BACKGROUND: Resilience, the ability to adapt and thrive under adversity, is essential for nursing students vulnerable to stress, burnout, and psychological distress. Strengthening resilience may protect well-being and enhance patient care.

METHODS: This feasibility study employed a one-group pretest-posttest design. A face-to-face stress management and resilience training (SMART) programme was delivered to 40 nursing undergraduates at a self-financing institution in Hong Kong. Resilience and stress were measured at baseline (T0), 2 weeks (T1), and 3 months (T2) using the Connor-Davidson Resilience Scale-10 (CD-RISC-10) and Perceived Stress Scale-10 (PSS-10). A post-training satisfaction survey was also conducted.

RESULTS: Thirty-one participants completed all assessments. CD-RISC-10 scores increased significantly from 22.00 (SD = 4.11) at T0 to 27.81 (SD = 7.39) at T1 and 26.42 (SD = 6.45) at T2. Improvements were significant between T0-T1 (p < 0.001) and T0-T2 (p = 0.004). PSS-10 scores declined slightly (18.61 at T0 to 17.65 at T2) but without statistical significance (p = 0.123). Satisfaction ratings were high (mean 5.38-5.60/6), particularly for trainer performance and content.

CONCLUSION: The SMART programme was feasible, well-received, and effective in improving resilience among Hong Kong nursing undergraduates. Findings support its potential integration into undergraduate nursing curricula.

PMID:41268651 | DOI:10.1111/jcap.70045

Biometrics. 2025 Oct 8;81(4):ujaf146. doi: 10.1093/biomtc/ujaf146.

ABSTRACT

Motivated by a malaria vaccine efficacy trial, this paper investigates generalized nonparametric temporal models of intensity processes with multiple time scales. Through the choice of link functions, the proposed models encompass a wide range of models such as the multiplicative temporal intensity model and the additive temporal intensity model. A maximum likelihood estimation procedure is developed to estimate the effects of two time-scales via the local linear smoothing with double kernels. Computational algorithms are developed to facilitate applications of the proposed method. An adaptive algorithm is developed to overcome the challenges of overlapping covariates. A cross-validation bandwidth selection procedure based on the logarithm of likelihood criteria is discussed. The asymptotic properties of the proposed estimators are investigated. Our simulation study shows that the proposed methods have satisfactory finite sample performance for both the multiplicative temporal intensity model and additive temporal intensity model. The proposed methods are applied to analyze the MAL-094/MAL-095 malaria vaccine efficacy trial data to investigate how the new malaria infection risk changes over time and how a prior infection or vaccination changes the future infection risk. The proposed method provides new insight into the protective effects of the malaria vaccine against new malaria infections and how the vaccine efficacy is modified by the history of prior malaria infection over time.

PMID:41268646 | DOI:10.1093/biomtc/ujaf146

Biometrics. 2025 Oct 8;81(4):ujaf150. doi: 10.1093/biomtc/ujaf150.

ABSTRACT

Federated learning enables the training of a global model while keeping data localized; however, current methods face challenges with high-dimensional semiparametric models that involve complex nuisance parameters. This paper proposes a federated double machine learning framework designed to address high-dimensional nuisance parameters of semiparametric models in multicenter studies. Our approach leverages double machine learning (Chernozhukov et al., 2018a) to estimate center-specific parameters, extends the surrogate efficient score method within a Neyman-orthogonal framework, and applies density ratio tilting to create a federated estimator that combines local individual-level data with summary statistics from other centers. This methodology mitigates regularization bias and overfitting in high-dimensional nuisance parameter estimation. We establish the estimator’s limiting distribution under minimal assumptions, validate its performance through extensive simulations, and demonstrate its effectiveness in analyzing multiphase data from the Alzheimer’s Disease Neuroimaging Initiative study.

PMID:41268645 | DOI:10.1093/biomtc/ujaf150