Category: Statistics

J Breast Imaging. 2026 Feb 9:wbaf044. doi: 10.1093/jbi/wbaf044. Online ahead of print.

ABSTRACT

OBJECTIVE: To determine the contrast-enhanced mammography-guided biopsy (CEM-Bx) success rate for MRI-suspicious lesions lacking known tomosynthetic or sonographic correlate along with factors associated with biopsy success.

METHODS: From June 2022 to August 2023, this prospective IRB-approved study enrolled women with breast MRI lesions rated BI-RADS ≥4A for CEM-Bx. Ipsilateral contrast-enhanced mammography (CEM) was performed in the biopsy suite and correlated with MRI. For visible lesions, CEM-Bx was performed immediately after prebiopsy CEM. Success criteria included enhancing correlate visualization and biopsy completion with accurate lesion sampling. An MRI-guided biopsy was recommended for failures. The success rate was evaluated with the Wilson score interval. MRI lesion and patient characteristics (size, type [mass, focus, or nonmass enhancement], kinetics, breast density, body mass index, background parenchymal enhancement [BPE], radiologist CEM experience, radiologist MRI experience, and histopathology) were collected. Multivariable logistic regression was performed with backward feature selection.

RESULTS: Analysis included 152 women (mean age 53 ± 11 years) with 184 lesions. CEM-Bx was successful for 106/184 (57.6%; [95% CI, 50.0-65.0]) lesions with 24/106 (22.6%) malignant. Of 78 failures, 60 (76.9%) lacked enhancement on prebiopsy CEM, 14 (17.9%) were not visualized with the biopsy grid, and 4 (5.1%) were not accurately sampled; 14/78 (17.9%) failures proved malignant. Increasing lesion size (odds ratio [OR] = 1.03; [95% CI, 1.01-1.06]), more years of radiologist CEM experience (OR = 1.24; [95% CI, 1.01-1.49]), and lower BPE (OR = 0.68 [95% CI, 0.46-0.98]) were associated with success.

CONCLUSION: Contrast-enhanced mammography biopsy can be a successful alternative to MRI-guided biopsy for MRI-detected lesions. MRI-guided biopsy should be pursued if CEM-Bx fails.

PMID:41661663 | DOI:10.1093/jbi/wbaf044

J Med Internet Res. 2026 Feb 9;28:e78233. doi: 10.2196/78233.

ABSTRACT

BACKGROUND: The rapid proliferation of electronic devices has increased screen time, raising concerns about its potential health effects, including chronic pain. However, existing studies have limitations in scope and causal inference, with inconsistent findings and a lack of exploration of potential biological mechanisms.

OBJECTIVE: The objective of our study was to investigate the causal associations and potential shared biological mechanisms between different forms of screen time and various chronic pain phenotypes.

METHODS: Leveraging genome-wide association study data, we investigated the association and potential shared biological mechanisms between screen time (time spent watching television, time spent using computer, and length of mobile phone use) and chronic pain phenotypes (including multisite chronic pain [MCP], back, knee, neck or shoulder, hip pain, and headaches). Two-sample Mendelian randomization (MR), reverse MR and multivariable Mendelian randomization (MVMR) analysis were performed to examine associations between screen time and chronic pain. Summary data-based Mendelian randomization (SMR), transcriptome-wide association study (TWAS), and colocalization analysis were used to identify the shared genes and potential biological mechanism.

RESULTS: MR analysis revealed that time spent watching television and length of mobile phone use were positively associated with several types of chronic pain, while time spent using computer showed a negative association. Specifically, time spent watching television was positively associated with the risk of MCP (P=1.05×10^-31; odds ratio [OR] 1.61, 95% CI 1.49-1.74), back pain (P=2.41×10^-8; OR 1.14, 95% CI 1.09-1.19), knee pain (P=7.10×10^-6; OR 1.09, 95% CI 1.05-1.13), neck or shoulder pain, and hip pain. Length of mobile phone use was positively associated with the risk of MCP (P=2.15×10^-5; OR 1.22, 95% CI 1.11-1.34), headaches, and neck or shoulder pain. However, time spent using computer was negatively associated with the risk of MCP (P<.001; OR 0.83, 95% CI 0.75-0.92), back pain, and knee pain. The reverse MR results showed that MCP was positively associated with time spent watching television (P=4.8×10^-7; OR 1.27, 95% CI 1.16-1.4) and length of mobile phone use (P=3.38×10^-5; OR 1.29, 95% CI 1.14-1.45), while the association with time spent using computer (P=.61; OR 0.97, 95% CI 0.87-1.09) was not statistically significant. The MVMR results failed to meet the criterion that all conditional F-statistics exceed 10. Integrative 3 analysis methods identified overlapping genes, with CEP170 emerging as a key gene consistently supported by SMR, TWAS, and colocalization analysis in the relationship between time spent using computer and MCP.

CONCLUSIONS: Our findings demonstrate an association between screen time and various aspects of chronic pain. The CEP170 gene might contribute to the shared biological mechanism between time spent using computer and MCP risk. However, due to the absence of robust MVMR results, the potential influence of confounding factors cannot be ruled out.

PMID:41661662 | DOI:10.2196/78233

JAMA Pediatr. 2026 Feb 9. doi: 10.1001/jamapediatrics.2025.6120. Online ahead of print.

ABSTRACT

IMPORTANCE: Autism spectrum disorder (ASD), intellectual disability (ID), and cerebral palsy (CP) are lifelong neurodevelopmental conditions accompanied by varying impairments. US mortality data for these groups are limited.

OBJECTIVE: To compare mortality and causes of death among a multisite cohort identified at age 8 years with ASD, ID, or CP with the general population through youth or young adulthood.

DESIGN, SETTING, AND PARTICIPANTS: Nine US sites identified 32 787 individuals who met case definitions for ASD, ID, and/or CP at age 8 years during active population-based cross-sectional surveillance conducted biennially from 2000 through 2016 by the US Centers for Disease Control and Prevention’s Autism and Developmental Disabilities Monitoring (ADDM) Network. Individuals were linked to death certificates through 2021. Cases with multiple conditions (18.9%) were included in each case group. General population data from the National Vital Statistics System were matched to ADDM Network sites and years of participation. Analyses were completed in 2024.

EXPOSURE(S): ASD, ID, or CP.

MAIN OUTCOMES AND MEASURES: Death and International Classification of Diseases, 10th revision (ICD-10) International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) causes from death certificate linkage.

RESULTS: There were 145 deaths among 23 393 people with ASD, 285 deaths among 14 031 people with ID, and 123 deaths among 1612 people with CP. Increased mortality compared with the general population was seen for ASD (hazard ratio [HR], 1.35; 95% CI, 1.15-1.59), ID (HR, 4.35; 95% CI, 3.87-4.88), and CP (HR, 9.62; 95% CI, 8.06-11.48). Further stratified by sex and co-occurring ID, mortality for ASD was higher only for females with co-occurring ID (HR, 5.04; 95% CI,3.21-7.91) compared with females in the general population. The distribution of causes of death varied across groups. The most common underlying cause of death ICD-10 chapters were external causes of morbidity and mortality (V01-Y98) for the general population and ASD case group, and diseases of the nervous system (G00-G99) for CP and ID case groups. The only ICD-10 chapter hazard of death that was not elevated for ID and CP compared with the general population was external causes as underlying cause of death. Mortality from external causes was also not elevated as underlying or any cause of death for ASD. There were also notable subchapter mortality differences with important clinical and public health implications. Only 11% of those with ASD, 1% of those with ID, and 49% of those with CP had an ICD-10 code for the respective disability on their death certificate.

CONCLUSIONS AND RELEVANCE: In this study, individuals with ASD, ID, or CP experienced higher mortality from a range of causes compared with the general population in youth and young adulthood. Mortality among these groups is difficult to ascertain using death certificates alone, since ICD-10 codes for these disabilities were rarely listed. These findings can inform public health and health care strategies to understand and prevent health disparities and excess mortality associated with developmental disabilities.

PMID:41661606 | DOI:10.1001/jamapediatrics.2025.6120

JAMA. 2026 Feb 9. doi: 10.1001/jama.2025.27259. Online ahead of print.

ABSTRACT

IMPORTANCE: Evidence linking coffee and tea to cognitive health remains inconclusive, and most studies fail to differentiate caffeinated from decaffeinated coffee.

OBJECTIVE: To investigate associations of coffee and tea intake with dementia risk and cognitive function.

DESIGN, SETTING, AND PARTICIPANTS: Prospective cohort study that included female participants from the Nurses’ Health Study (NHS; n = 86 606 with data from 1980-2023) and male participants from the Health Professionals Follow-up Study (HPFS; n = 45 215 with data from 1986-2023) who did not have cancer, Parkinson disease, or dementia at study entry (baseline) in the US.

EXPOSURES: The primary exposures were intakes of caffeinated coffee, decaffeinated coffee, and tea. Dietary intake was collected every 2 to 4 years using validated food frequency questionnaires.

MAIN OUTCOMES AND MEASURES: The primary outcome was dementia, which was identified via death records and physician diagnoses. The secondary outcomes included subjective cognitive decline assessed by a questionnaire-based score (range, 0-7; higher scores indicate greater perceived decline; cases defined as those with a score ≥3) and objective cognitive function assessed only in the NHS cohort using telephone-based neuropsychological tests such as the Telephone Interview for Cognitive Status (TICS) score (range, 0-41) and a measure of global cognition (a standardized mean z score for all 6 administered cognitive tests).

RESULTS: Among 131 821 participants (mean age at baseline, 46.2 [SD, 7.2] years in the NHS cohort and 53.8 [SD, 9.7] years in the HPFS cohort; 65.7% were female) during up to 43 years of follow-up (median, 36.8 years; IQR, 28-42 years), there were 11 033 cases of incident dementia. After adjusting for potential confounders and pooling results across cohorts, higher caffeinated coffee intake was significantly associated with lower dementia risk (141 vs 330 cases per 100 000 person-years comparing the fourth [highest] quartile of consumption with the first [lowest] quartile; hazard ratio, 0.82 [95% CI, 0.76 to 0.89]) and lower prevalence of subjective cognitive decline (7.8% vs 9.5%, respectively; prevalence ratio, 0.85 [95% CI, 0.78 to 0.93]). In the NHS cohort, higher caffeinated coffee intake was also associated with better objective cognitive performance. Compared with participants in the lowest quartile, those in the highest quartile had a higher mean TICS score (mean difference, 0.11 [95% CI, 0.01 to 0.21]) and a higher mean global cognition score (mean difference, 0.02 [95% CI, -0.01 to 0.04]); however, the association with global cognition was not statistically significant (P = .06). Higher intake of tea showed similar associations with these cognitive outcomes, whereas decaffeinated coffee intake was not associated with lower dementia risk or better cognitive performance. A dose-response analysis showed nonlinear inverse associations of caffeinated coffee and tea intake levels with dementia risk and subjective cognitive decline. The most pronounced associated differences were observed with intake of approximately 2 to 3 cups per day of caffeinated coffee or 1 to 2 cups per day of tea.

CONCLUSIONS AND RELEVANCE: Greater consumption of caffeinated coffee and tea was associated with lower risk of dementia and modestly better cognitive function, with the most pronounced association at moderate intake levels.

PMID:41661604 | DOI:10.1001/jama.2025.27259

JAMA Netw Open. 2026 Feb 2;9(2):e2557241. doi: 10.1001/jamanetworkopen.2025.57241.

ABSTRACT

IMPORTANCE: Rates of posttraumatic stress disorder (PTSD) and major depressive disorder (MDD) are high among paramedics.

OBJECTIVE: To evaluate the efficacy of a cognitive resilience training program for reducing the development of PTSD and MDD among early career paramedics compared with psychoeducation and standard practice.

DESIGN, SETTING, AND PARTICIPANTS: This randomized clinical trial of paramedics training at 15 universities across England was conducted between October 2017 to October 2022 with 12-month follow-up. Data were analyzed from December 2023 to July 2025.

INTERVENTION: Participants were randomized to receive internet-delivered cognitive training in resilience (iCT-R), psychoeducation, or standard practice. iCT-R, a guided online intervention that utilizes cognitive therapy tools to target predictors of PTSD and MDD identified in prospective research with paramedics, consisted of 6 modules delivered over 6 weeks with 6-monthly top-up sessions delivered by email. Internet-delivered psychoeducation, a supported online psychoeducation intervention, consisted of 6 topics (1 topic per week) with 6-monthly top-up sessions delivered by email. Standard practice was training as usual.

MAIN OUTCOMES AND MEASURES: The primary outcome was rate of PTSD and MDD at 1-year follow-up, assessed by independent assessors blinded to intervention using the Structured Clinical Interview for the Diagnostic and Statistical Manual of Mental Disorders (Fifth Edition). Secondary outcomes included measures of PTSD and depression symptom severity, resilience, rumination, anxiety, psychological distress, and well-being. Intent-to-treat analyses were conducted, with the primary outcome analyzed using mixed-effects logistic regression.

RESULTS: Of 570 student paramedics enrolled (372 female [65.3%]; mean [SD] age, 23.67 [6.88] years), 195 were randomized to iCT-R, 197 to psychoeducation, and 178 to standard practice. For participants randomized to iCT-R, the odds of meeting criteria for PTSD or MDD at 12 months were significantly lower compared with psychoeducation (odds ratio [OR], 0.20; 95% CI, 0.05-0.73) and standard practice (OR, 0.25; 95% CI, 0.07-0.97). Providing iCT-R training to 18 to 24 paramedic trainees (number needed to treat) would prevent 1 case of PTSD or MDD.

CONCLUSIONS AND RELEVANCE: In this randomized clinical trial, participants receiving iCT-R were approximately 5 times less likely to develop PTSD or MDD at 1-year follow-up compared with psychoeducation and 4 times less likely when compared with standard practice. These findings suggest that iCT-R appears to decrease the likelihood of developing PTSD and MDD in early career paramedics.

TRIAL REGISTRATION: isrctn.org Identifier: ISRCTN16493616.

PMID:41661593 | DOI:10.1001/jamanetworkopen.2025.57241

Drugs Aging. 2026 Feb 9. doi: 10.1007/s40266-026-01280-2. Online ahead of print.

ABSTRACT

BACKGROUND: The aim of this study was to assess the prevalence of potentially inappropriate medication (PIM) use among adults ≥ 65 years, overall and for population subgroups, and factors associated with prevalent PIM use.

METHODS: Participant and medications data from the Atherosclerosis Risk in Communities (ARIC), Multi-ethnic Study of Atherosclerosis (MESA), and Action for Health in Diabetes (Look AHEAD) were used. The total number of individuals contributing to analysis was 9439 for ARIC, 5223 for MESA, and 3771 for Look AHEAD. Participants’ medication data were collected at study exams with medication inventories. Prevalence of any PIM use (yes/no), total number of PIMs used, and the major drug classes of PIMs used within a cohort were identified using Beers Criteria closest to the time of study exam (1997 onward) for all participants aged 65 years or older. Multivariable adjusted logistic regression was used to assess demographic and clinical factors associated with prevalence of any Beers Criteria PIM at the first exam after turning 65 years of age, separately for each cohort.

RESULTS: The prevalence of PIM use at the first exam at ≥ 65 years was 67% in ARIC, 51% in MESA, and 70% in Look AHEAD. The most prevalently used PIM classes across cohorts were non-aspirin pain medications, proton-pump inhibitors, and sulfonylureas. Higher body mass index and diabetes were consistently associated with greater odds of PIM use across cohorts.

CONCLUSIONS: Use of potentially inappropriate prescription and nonprescription medications was highly prevalent across three diverse cohorts and highlights the need for clinicians and pharmacists to review patient medication lists and optimize management of medications.

PMID:41661525 | DOI:10.1007/s40266-026-01280-2

Infection. 2026 Feb 9. doi: 10.1007/s15010-026-02739-5. Online ahead of print.

ABSTRACT

BACKGROUND: Multidrug-resistant Gram-negative bacteria (GNB), including carbapenem-resistant Enterobacterales and Pseudomonas aeruginosa, pose a growing global health threat with limited treatment options. Imipenem-relebactam (IMI/REL) is a promising therapy, but emerging resistance patterns remain poorly defined worldwide.

OBJECTIVES: This meta-analysis aimed to provide a comprehensive assessment of global IMI/REL resistance among Enterobacterales and non-fermenting GNB, highlighting species-specific, geographic, and temporal patterns.

METHODS: Studies were searched in Scopus, PubMed, and EMBASE (until October 23 2024), and all statistical analyses were conducted using R (ver. 4.2.1).

RESULTS: A total of 149,396 Enterobacterales and non-fermenting GNB were included. Overall, IMI/REL resistance was low at 8.8% (95% CI 7.3-10.5), with Enterobacterales showing the lowest resistance (2.9%) and Pseudomonas the highest (30.7%). Resistance was higher in carbapenemase-producing Enterobacterales and non-fermenting species, and varied by infection source and geography. Temporal analysis indicated a rising trend in resistance over recent years, while data were mostly derived from the Americas, limiting global generalizability.

CONCLUSION: IMI/REL remains largely effective against Enterobacterales, including ESBL and MDR strains, but resistance is increasing in high-risk subgroups and non-fermenting bacteria. These findings underscore the need for local susceptibility testing, cautious empiric therapy, and robust antimicrobial stewardship to preserve the efficacy of IMI/REL.

PMID:41661521 | DOI:10.1007/s15010-026-02739-5

Dermatol Ther (Heidelb). 2026 Feb 9. doi: 10.1007/s13555-026-01663-8. Online ahead of print.

ABSTRACT

INTRODUCTION: Hand eczema and psoriasis present overlapping clinical features, complicating diagnosis and treatment selection. Traditional diagnostic methods rely on clinical assessment, patient history, allergy testing, and histopathology. However, recent advancements in molecular diagnostics offer promising alternatives. Accurate diagnosis is crucial for optimal therapy selection, particularly in occupational dermatology, where hand eczema is a common occupational disease.This study aims to assess the effectiveness of molecular diagnostics in distinguishing hand eczema from psoriasis, analyze disease severity and chronicity, and evaluate therapeutic changes and health-related quality of life (HrQoL) over a 2-year period.

METHODS: A long-term cohort study was initiated in November 2020, enrolling 287 patients with suspected occupational skin disease. Molecular classification based on gene expression (CCL27 and NOS2) was performed on skin biopsies. Data collection included physician global assessment (PGA), Quality of Life in Hand Eczema Questionnaire (QOLHEQ), and Dermatology Life Quality Index (DLQI). Statistical analyses employed Cohen’s kappa, χ² tests, Wilcoxon signed-rank tests, and 95% confidence intervals.

RESULTS: Of 272 patients assessed via molecular diagnostics, 38.9% had clinically unclear diagnoses, with over 95% of these clarified through molecular classification. Dermatological and molecular diagnoses showed low agreement. Disease severity and chronicity significantly decreased over 2 years. Use of systemic therapies increased, while overall corticosteroid usage declined. HrQoL improved significantly, with DLQI scores decreasing by 50%.

CONCLUSIONS: Molecular diagnostics significantly enhance diagnostic accuracy for hand dermatoses, leading to targeted treatment adjustments. The observed therapy shift correlated with improved disease outcomes and HrQoL. As specialized systemic treatments emerge, precise diagnostic tools will be essential for optimizing patient care and reducing the burden of occupational skin diseases.

PMID:41661519 | DOI:10.1007/s13555-026-01663-8

Dig Dis Sci. 2026 Feb 9. doi: 10.1007/s10620-026-09715-x. Online ahead of print.

ABSTRACT

BACKGROUND: Endoscopic full-thickness resection (EFTR) can effectively remove extraluminal tumors, overcoming limitations of endoscopic submucosal dissection (ESD). However, optimal post-EFTR feeding timing lacks standardized guidelines for patients with gastric tumors. This study aims to assess the safety of early feeding after EFTR.

METHODS: Retrospective analysis was conducted on patients who underwent EFTR at our hospital between January 2014 and January 2019. Based on actual fasting duration, patients were categorized: short fasting (≤ 2 days, n = 72) and long fasting (> 2 days, n = 431) group. Using 1:1 propensity score matching, postoperative complications and hospital stay were compared between balanced groups.

RESULTS: A total of 503 patients were included in this study, among which 72 were in the short fasting group and 431 were in the long fasting group. After matching, the baseline characteristics of 68 patients in the short fasting group and 68 patients in the long fasting group reached equilibrium (P > 0.05). The average age was 53.82 ± 10.98 years old. There was no significant difference in clinicopathological conditions or lesion size between the two groups. There was no statistically significant difference in the rates of postoperative bleeding, fever and abdominal distension between the two groups of patients after EFTR. However, compared with the long fasting protocol, a trend of shorter hospitalization was observed in the short fasting group.

CONCLUSIONS: Compared with the long fasting protocol, early feeding after EFTR for gastric tumors did not increase the incidence of discomfort or postoperative complications. In addition, short fasting protocol has a tendency to shorten hospital stays, which represents potential clinical benefits.

PMID:41661495 | DOI:10.1007/s10620-026-09715-x

Eat Weight Disord. 2026 Feb 9. doi: 10.1007/s40519-026-01821-z. Online ahead of print.

ABSTRACT

PURPOSE: The relationship between dietary nutrient intake and metabolically healthy obesity (MHO) remains poorly understood. This study aimed to construct machine learning models to predict MHO based on dietary nutrient profiles and to identify the most influential nutrients contributing to this phenotype.

METHODS: Data were derived from the U.S. National Health and Nutrition Examination Survey (NHANES) 2005-2018. Forty-five dietary nutrients, along with demographic and lifestyle variables, were included in two predictive frameworks: a dietary-only model and a complete model. Feature preprocessing involved assessing mixture effects, removing multicollinear variables, addressing class imbalance, and selecting important predictors. Six machine learning algorithms-random forest (RF), light gradient-boosting machine, k-nearest neighbor, Naive Bayes, support vector machine, and eXtreme Gradient Boosting (XGBoost)-were developed and benchmarked to compare performance. Model interpretability was examined using SHapley Additive exPlanations (SHAP) and Local Interpretable Model-agnostic Explanations (LIME).

RESULTS: A total of 8914 participants, including 475 classified as having MHO, were analyzed. The Random Forest model exhibited the best predictive performance in the complete model, achieving training and validation AUCs of 0.986 and 0.991, respectively. In contrast, XGBoost demonstrated superior performance in the dietary-only model, with AUCs of 0.971 and 0.988. SHAP and LIME analyses revealed that added vitamin B12, lycopene, caffeine, theobromine, and lutein/zeaxanthin were the strongest positive predictors in the complete model. When only dietary factors were considered, lycopene, lutein/zeaxanthin, magnesium, potassium, and selenium emerged as the most influential nutrients.

CONCLUSIONS: RF and XGBoost models provided the highest predictive accuracy for MHO using complete and dietary feature sets, respectively. The consistent findings from SHAP and LIME analyses emphasized lycopene and lutein/zeaxanthin as reliable and biologically relevant key predictors of metabolically healthy obesity.

LEVEL OF EVIDENCE: Level III, well-designed cohort or case-control analytic study.

PMID:41661492 | DOI:10.1007/s40519-026-01821-z