Category: Statistics

JAMA Neurol. 2025 Feb 24. doi: 10.1001/jamaneurol.2024.5406. Online ahead of print.

ABSTRACT

IMPORTANCE: A leading cause of surgically remediable, drug-resistant focal epilepsy is focal cortical dysplasia (FCD). FCD is challenging to visualize and often considered magnetic resonance imaging (MRI) negative. Existing automated methods for FCD detection are limited by high numbers of false-positive predictions, hampering their clinical utility.

OBJECTIVE: To evaluate the efficacy and interpretability of graph neural networks in automatically detecting FCD lesions on MRI scans.

DESIGN, SETTING, AND PARTICIPANTS: In this multicenter diagnostic study, retrospective MRI data were collated from 23 epilepsy centers worldwide between 2018 and 2022, as part of the Multicenter Epilepsy Lesion Detection (MELD) Project, and analyzed in 2023. Data from 20 centers were split equally into training and testing cohorts, with data from 3 centers withheld for site-independent testing. A graph neural network (MELD Graph) was trained to identify FCD on surface-based features. Network performance was compared with an existing algorithm. Feature analysis, saliencies, and confidence scores were used to interpret network predictions. In total, 34 surface-based MRI features and manual lesion masks were collated from participants, 703 patients with FCD-related epilepsy and 482 controls, and 57 participants were excluded during MRI quality control.

MAIN OUTCOMES AND MEASURES: Sensitivity, specificity, and positive predictive value (PPV) of automatically identified lesions.

RESULTS: In the test dataset, the MELD Graph had a sensitivity of 81.6% in histopathologically confirmed patients seizure-free 1 year after surgery and 63.7% in MRI-negative patients with FCD. The PPV of putative lesions from the 260 patients in the test dataset (125 female [48%] and 135 male [52%]; mean age, 18.0 [IQR, 11.0-29.0] years) was 67% (70% sensitivity; 60% specificity), compared with 39% (67% sensitivity; 54% specificity) using an existing baseline algorithm. In the independent test cohort (116 patients; 62 female [53%] and 54 male [47%]; mean age, 22.5 [IQR, 13.5-27.5] years), the PPV was 76% (72% sensitivity; 56% specificity), compared with 46% (77% sensitivity; 47% specificity) using the baseline algorithm. Interpretable reports characterize lesion location, size, confidence, and salient features.

CONCLUSIONS AND RELEVANCE: In this study, the MELD Graph represented a state-of-the-art, openly available, and interpretable tool for FCD detection on MRI scans with significant improvements in PPV. Its clinical implementation holds promise for early diagnosis and improved management of focal epilepsy, potentially leading to better patient outcomes.

PMID:39992650 | DOI:10.1001/jamaneurol.2024.5406

J Mater Chem B. 2025 Feb 24. doi: 10.1039/d4tb02835j. Online ahead of print.

ABSTRACT

Novel and promising biomaterials for bone tissue engineering have been investigated over the years. Aiming to contribute to this progress, this study developed and evaluated polycaprolactone (PCL) scaffolds with 5% (w/w) 58S-bioactive glass (58S-BG) fabricated via melt electrowriting (MEW). Morphological and chemical characterization of the scaffolds was conducted. The biological potential was assessed in vitro with alveolar bone-derived mesenchymal stem cells through cytotoxicity, adhesion, protein production, alkaline phosphatase activity, and mineral nodule formation assays. In vivo, scaffolds implanted in rats were analyzed for biocompatibility, inflammation, and degradation using H&E staining and immunohistochemical markers for angiogenesis and macrophage polarization. Statistical analysis was performed at a 5% significance level. Appropriate fiber alignment but a higher fiber diameter was found for PCL + BG5% compared to PCL scaffolds (p = 0.002). EDS spectra confirmed the presence of BG’s chemical components for BG-laden scaffolds, attesting to BG particle incorporation into the filaments. Raman spectroscopy evidenced the chemical nature of the BG powder, and FTIR spectra revealed -OH stretching for PCL + BG5%, evidencing its hydrophilic potential. None of the scaffolds were cytotoxic, and BG-laden formulation increased cell viability after 7 days (p = 0.0006), also showing greater cell adhesion/spreading over time compared to pristine PCL scaffolds. BG’s presence also increased the mineral matrix formation (p ≤ 0.0021) over 21 days and retained ALP activity after 14 days (p = 0.705) compared to PCL. In vivo, PCL scaffolds retained fiber alignment and preserved their volume throughout the evaluation, showing minimal structural alteration. In contrast, PCL + BG5% scaffolds showed more visible structural changes at 28 days. Despite this, the PCL + BG5% formulation remained biocompatible and significantly promoted angiogenesis compared to pristine PCL scaffolds. In sum, BG-laden scaffolds were successfully melt electrowritten, retaining the scaffolds’ porous architecture, showing appropriate properties, including cell viability, adhesion, mineralized nodule deposition, biocompatibility, and angiogenesis, indicating that these materials are a promising alternative for enhancing bone tissue regeneration.

PMID:39992649 | DOI:10.1039/d4tb02835j

JAMA Intern Med. 2025 Feb 24. doi: 10.1001/jamainternmed.2024.8129. Online ahead of print.

NO ABSTRACT

PMID:39992643 | DOI:10.1001/jamainternmed.2024.8129

JAMA Intern Med. 2025 Feb 24. doi: 10.1001/jamainternmed.2024.8432. Online ahead of print.

ABSTRACT

IMPORTANCE: National estimates of mortality among physicians and other health care workers are lacking. It is also unknown if distinct patterns exist across sex, race, and ethnicity.

OBJECTIVE: To compare all-cause and cause-specific mortality rates among physicians, health care workers, and non-health care workers by sex, race, and ethnicity.

DESIGN, SETTING, AND PARTICIPANTS: The National Vital Statistics System, a population-based registry of US death certificates, was used to obtain data on deaths among adults aged 25 to 74 years from January 2020 to December 2022 by usual occupation. Data were analyzed from January 2024 to December 2024.

EXPOSURES: Usual occupation, sex, race, and ethnicity.

MAIN OUTCOMES AND MEASURES: Overall and cause-specific mortality rates were calculated for each occupation, as well as sex, race, and ethnicity subgroups, and compared using mortality rate ratios. Mortality estimates were age-adjusted and sex-adjusted, and health care occupations were compared with non-health care occupations with similar income levels (categorized as low income, medium income, and high income based on US Census income terciles).

RESULTS: Most health care workers had lower age-adjusted and sex-adjusted annual mortality rates per 100 000 population than non-health care workers (eg, physicians [269.3], high-income non-health care workers [499.2], and non-health care workers overall [730.6]). While female individuals had lower mortality than male individuals in non-health care occupations overall (female to male ratio, 0.55; 95% CI, 0.55-0.55) and high-income non-health care occupations specifically (0.60; 95% CI, 0.60-0.60), this advantage was absent for several health care occupations, including physicians (0.97; 95% CI, 0.93-1.01). In particular, female physicians experienced higher mortality than male physicians of neoplasms and chronic lower respiratory diseases, despite lower mortality of these causes among female individuals in high-income non-health care occupations. Black workers had higher mortality than White workers across all occupations, although the Black to White mortality ratio was largest for physicians (2.13; 95% CI, 1.99-2.29), with the largest differences due to neoplasms, heart disease, and COVID-19. Black female physicians had higher mortality rates than all other physician subgroups and White female individuals in non-health care occupations. While Hispanic workers had lower mortality than White workers in non-health care occupations overall (Hispanic to White ratio, 0.83; 95% CI, 0.83-0.83) and high-income non-health care occupations specifically (0.90; 95% CI, 0.90-0.91), this pattern was reversed for several health care occupations, including physicians (1.18; 95% CI, 1.09-1.27).

CONCLUSIONS AND RELEVANCE: The results of this cross-sectional study suggest that although physicians and most health care workers experienced lower mortality rates compared with the general population, this benefit did not fully extend to female individuals or racial and ethnic minority groups. Renewed efforts are needed to address health inequities within the health care workforce.

PMID:39992637 | DOI:10.1001/jamainternmed.2024.8432

Probiotics Antimicrob Proteins. 2025 Feb 24. doi: 10.1007/s12602-025-10480-y. Online ahead of print.

ABSTRACT

Supplementation of probiotic strains can enhance the absorption of amino acids from protein in the gut. The purpose of this study was to assess if supplementation of a multi-strain probiotic or a postbiotic, consisting of the same strains, would alter the absorption of individual and total amino acids following ingestion of a plant-based meal. Sixteen male participants consumed either probiotic (PRO) or postbiotic (cells inactivated by γ-irradiation; POST), both consisting of L. paracasei LP-DG® (CNCM I-1572) plus L. paracasei LPC-S01 (DSM 26760), or a placebo (PLA) for 2 weeks in a randomized, double-blind, crossover design study separated by a 4-week washout period. During the testing session, blood samples were taken at baseline, 30-, 60-, 120-, and 180-min post-ingestion of a plant-based vegan burger patty. Plasma amino acid levels were analyzed, and percent changes from baseline were assessed using linear mixed-effects models, with the PLA condition as the reference group. There was statistically significant POST condition-by-time interactions for percent changes in alanine, asparagine, citrulline, cystine, glycine, methionine, proline, and total amino acids (p < 0.05, for all). Additionally, there was a statistically significant condition (PRO) by time interactions for cystine (p = 0.02). Two weeks of POST supplementation resulted in significant improvements in amino acid absorption profiles for various individual amino acids and total amino acids compared to PLA. This is the first study to report improved amino acid absorption from a mixed macronutrient meal following a period of postbiotic supplementation.

PMID:39992622 | DOI:10.1007/s12602-025-10480-y

Int J Health Econ Manag. 2025 Feb 24. doi: 10.1007/s10754-025-09392-9. Online ahead of print.

ABSTRACT

Baumol’s (Am Econ Rev 57: 415-426, 1967) model of ‘unbalanced growth’ yields a supply-side explanation for the ‘cost explosion’ in health care. Applying a testing strategy suggested by Hartwig (J Health Econ 27: 603-623, 2008), a sprawling literature affirms that the ‘Baumol effect’ has both a statistically and economically significant impact on health care expenditure growth. Skeptics maintain, however, that the proliferation of hi-tech medicine in acute care is clearly at odds with the assumption underlying Baumol’s model that productivity-enhancing machinery and equipment is only installed in the ‘progressive’ (i.e. manufacturing) sector of the economy. They argue that Baumol’s cost disease may affect long-term care, but not acute care. Our aim in this paper is to test whether Baumol’s cost disease affects long-term care and acute care differently. Our testing strategy consists in combining Extreme Bounds Analysis (EBA) with an outlier-robust MM estimator. Using panel data for 23 OECD countries, our results provide robust and statistically significant evidence that expenditures on both acute care and long-term care are driven by Baumol’s cost disease, even though the effect on long-term care expenditures is more pronounced.

PMID:39992610 | DOI:10.1007/s10754-025-09392-9

Gastric Cancer. 2025 Feb 24. doi: 10.1007/s10120-025-01598-7. Online ahead of print.

ABSTRACT

BACKGROUND: Single nucleotide polymorphisms (SNPs) are associated with various diseases, including gastric cancer. The ADAMTS14 gene is linked to multiple types of cancer. However, the relationship between ADAMTS14 and its genetic polymorphisms with susceptibility to gastric cancer (GC) and prognosis remains unclear.

METHODS: A case-control study was conducted involving 855 patients diagnosed with gastric cancer (GC) and an equal number of cancer-free controls. Following rigorous statistical analysis, molecular experiments were performed to elucidate the functional significance of the SNPs in the context of GC.

RESULTS: ADAMTS14 rs3740440 (OR = 1.45, p = 0.014) shows a significant association with increased GC risk, while rs11572 (OR = 0.42, p < 0.001) is associated with protection against GC. Moreover, patients with the (CG + GG) genotype of rs3740440 exhibit a poor prognosis (HR = 1.68, p = 0.007). Mechanistically, luciferase reporter assays revealed that the G allele of rs3740440 disrupts the binding of hsa-miR-4294 and hsa-miR-3198 to the 3′ untranslated region (3′ UTR) of ADAMTS14, leading to increased expression of ADAMTS14 and the promotion of malignant behaviors in GC cells.

CONCLUSIONS: Our findings underscore the significant role of ADAMTS14 SNPs in both the risk and prognosis of gastric cancer (GC), providing valuable insights into the underlying molecular mechanisms. Specifically, rs3740440 disrupts the interaction between ADAMTS14 and miRNA, resulting in increased expression of ADAMTS14. This heightened expression enhances its malignant biologic behaviors, indicating that rs3740440 could be a potential predictive marker for gastric cancer risk and prognosis.

PMID:39992602 | DOI:10.1007/s10120-025-01598-7

FASEB J. 2025 Feb 28;39(4):e70415. doi: 10.1096/fj.202402780R.

ABSTRACT

Inflammatory bowel disease (IBD) is a multifactorial disease, and patients frequently experience extraintestinal manifestations affecting multiple sites. Causes of systemic inflammation remain poorly understood, but molecules originating from the intestine likely play a role, with microbial and host small molecules polarizing host immune cells towards a pro- or anti-inflammatory phenotype. Using the dextran sodium sulfate (DSS) mouse model, which mimics the disrupted barrier function, microbial dysbiosis, and immune cell dysregulation of IBD, we investigated metabolomic and phenotypic changes at intestinal and systemic sites. Using spatial biology approaches, we mapped the distribution and relative abundance of molecules and cell types across a range of tissues, revealing significant changes in DSS-treated mice. Molecules identified as contributing to the statistical separation of treated from control mice were spatially localized within organs to determine their effects on cellular phenotypes through imaging mass cytometry. This spatial approach identified both intestinal and systemic molecular drivers of inflammation, including several not previously implicated in inflammation linked to IBD or the systemic effects of intestinal inflammation. Metabolic and inflammatory pathway interplay underpins systemic disease, and determining drivers at the molecular level may aid the development of new targeted therapies.

PMID:39989432 | DOI:10.1096/fj.202402780R

Eur J Breast Health. 2025 Feb 24. doi: 10.4274/ejbh.galenos.2025.2024-12-9. Online ahead of print.

ABSTRACT

OBJECTIVE: Breast cancer (BC) is a global concern due to its high incidence worldwide. The alarming increase in BC cases highlights the need for careful management of the disease at multiple levels. This study investigated the diagnostic value of hemoglobin, albumin, lymphocyte and platelet counts (HALP score), neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR) and lymphocyte-to-monocyte ratio (LMR) in newly diagnosed BC patients.

MATERIALS AND METHODS: A total of 84 individuals, including 42 healthy volunteers (group I) and 42 patients newly diagnosed with BC (group II), were included. Serum albumin levels were determined using spectrophotometry. The levels of tumor-markers carcinoembryonic antigen (CEA) and cancer antigen 15-3 (CA 15-3) in serum were analyzed by electrochemiluminescence immunoassay. Hemogram parameters were analyzed using fluorescence flow cytometry.

RESULTS: The median PLR was significantly lower in group II than group I (p = 0.014). There were no statistical differences in HALP score, NLR, LMR, and prognostic nutrition index between the two groups (p = 0.133, p = 0.993, p = 0.591, and p = 0.294, respectively). The sensitivity and specificity of PLR in predicting BC were 61.90% and 64.29%, respectively, with an area under the curve of 0.665 (p = 0.009, 95% confidence interval: 0.5480 to 0.7819, cut-off value ≤124). PLR, CEA and CA 15-3 were independent risk factors for BC (p<0.05).

CONCLUSION: The findings suggest that PLR may serve as a potential biomarker for the early diagnosis of BC; however, further validation is required. Conversely, the HALP score and other parameters did not demonstrate a significant association with early BC diagnosis. These results warrant corroboration through regional and community-based studies.

PMID:39989379 | DOI:10.4274/ejbh.galenos.2025.2024-12-9

Eur J Psychotraumatol. 2025 Dec;16(1):2465082. doi: 10.1080/20008066.2025.2465082. Epub 2025 Feb 24.

ABSTRACT

Objective: Trauma victimization is common among children, however, a significant proportion of trauma victims go unrecognized unless they are thoroughly assessed, even in child psychiatry clinics. The aim of this study was to evaluate the psychometric properties and diagnostic accuracy of the Turkish version of the UCLA PTSD Reaction Index for DSM-5 (PTSD RI-5) in a clinical sample of trauma-exposed children and adolescents.Method: A total of 208 children and adolescents admitted to the child psychiatry clinic, each of whom had a history of at least one traumatic event, were evaluated with the PTSD RI-5 to investigate trauma history and PTSD symptoms. All participants also completed the Revised Child Anxiety and Depression Scale (RCADS) and 64 participants were assessed with a semi-structured diagnostic interview for PTSD and depression.Results: Internal consistency for the total scale was high (Cronbach’s α = 0.91) and the confirmatory factor analysis (CFA) supported the four-factor structure of the PTSD RI-5 (CFI = 0.915, TLI = 0.902, RMSEA =0.062). ROC analysis showed strong diagnostic accuracy (AUC = 0.94).Conclusion: The Turkish version of the PTSD RI-5 may a reliable and valid tool for diagnosing PTSD in clinical samples and may improve diagnosis and treatment outcomes by identifying unrecognized trauma-related symptoms.

PMID:39989342 | DOI:10.1080/20008066.2025.2465082