Our study aimed to determine the prevalence and clinical phenotypes of JAK2 pathogenic mutation carriers in the CNSR-III ischemic stroke (IS) cohort, and to develop a pre-test genetic screening model for identifying high-risk individuals. We performed retrospective characterization of JAK2 pathogenic variants using targeted sequencing data in the CNSR-III cohort. Clinical and laboratory characteristics of JAK2 V617F mutation carriers and non-carriers were tested in a logistic regression model to identify key features. V617F screening score was developed to predict positive JAK2 V617F test results. 46 cases (0.4

Low-field portable magnetic resonance imaging (pMRI) devices address a crucial requirement in the realm of healthcare by offering the capability for on-demand and timely access to MRI, especially in the context of routine stroke emergency. Nevertheless, images acquired by these devices often exhibit poor clarity and low resolution, resulting in their reduced potential to support precise diagnostic evaluations and lesion quantification. In this paper, we propose a 3D deep learning based model, named Stroke-Aware CycleGAN (SA-CycleGAN), to enhance the quality of low-field images for further improving diagnosis of routine stroke. Firstly, based on traditional CycleGAN, SA-CycleGAN incorporates a prior of stroke lesions by applying a novel spatial feature transform mechanism. Secondly, gradient difference losses are combined to deal with the problem that the synthesized images tend to be overly smooth. We present a dataset comprising 101 paired high-field and low-field diffusion-weighted imaging (DWI), which were acquired through dual scans of the same patient in close temporal proximity. Our experiments demonstrate that SA-CycleGAN is capable of generating images with higher quality and greater clarity compared to the original low-field DWI. Additionally, in terms of quantifying stroke lesions, SA-CycleGAN outperforms existing methods. The lesion volume exhibits a strong correlation between the generated images and the high-field images, with R=0.852. In contrast, the lesion volume correlation between the low-field images and the high-field images is notably lower, with R=0.462. Furthermore, the mean absolute difference in lesion volumes between the generated images and high-field images ( $1.73\pm 2.03$ mL) was significantly smaller than the difference between the low-field images and high-field images ( $2.53\pm 4.24$ mL). It shows that the synthesized images not only exhibit superior visual clarity compared to the low-field acquired images, but also possess a high degree of consistency with high-field images. In routine clinical practice, the proposed SA-CycleGAN offers an accessible and cost-effective means of rapidly obtaining higher-quality images, holding the potential to enhance the efficiency and accuracy of stroke diagnosis in routine clinical settings. The code and trained models will be released on GitHub: SA-CycleGAN.

OBJECTIVE:The objective of this study was to compare clinical features and prognosis of late-onset neuromyelitis optica spectrum disorder (LO-NMOSD, onset age ≥60 years) with adult-onset NMOSD (AO-NMOSD, onset age 18-59 years), and to provide insights for individualized management in elderly patients. METHODS:Data from 748 patients with NMOSD (diagnosed according to the 2015 International Panel for NMO Diagnosis criteria) in the China National Registry of Neuro-Inflammatory Diseases (CNRID) were analyzed. Patients were stratified into AO-NMOSD (18-59 years, n = 617) and LO-NMOSD (≥ 60 years, n = 131). Demographics, clinical manifestations, imaging, treatments, and outcomes were compared using appropriate statistical methods including Kaplan-Meier survival curves and Cox proportional hazards regression. RESULTS:LO-NMOSD showed distinct traits: a lower female predominance (76.34% vs 86.55%), higher transverse myelitis (TM) incidence at onset (57.36% vs 40.17%), elevated annualized relapse rate (ARR; 0.52 ± 0.03 vs 0.38 ± 0.01), and accelerated disability (median Expanded Disability Status Scale [EDSS] 4.75 vs 3.0). TM-predominant relapses (39 of 45, 86.67% in LO vs 96 of 148, 64.86% in AO) contributed significantly to disability. Kaplan-Meier analysis showed LO-NMOSD had a higher risk of relapse (hazard ratio [HR] = 1.932, 95% confidence interval [CI] = 1.427-2.615), disability (HR = 3.192, 95% CI = 1.932-5.274) and reaching visual acuity (VA) ≤20 of 30 (HR = 3.523, 95% CI = 1.585-7.828). Cox regression confirmed that onset age ≥60 years was an independent risk factor for relapse (HR = 2.05, 95% CI = 1.60-2.59), disability (HR = 3.16, 95% CI = 2.14-4.62), and reaching VA ≤20 of 30 (HR 3.26, 95% CI = 1.83-5.48). INTERPRETATION:LO-NMOSD is characterized by myelitis-predominance with recurrent spinal cord involvement, high risk of relapses, and severe disability. It thus underscores the need for heightened clinical attention, with rigorous monitoring that balance safety and efficacy for elderly patients with NMOSD. ANN NEUROL 2026;99:1139-1151.

BACKGROUND:Post-stroke cognitive impairment (PSCI) is a frequent and disabling consequence of ischemic stroke. Leukocyte telomere length (LTL), a biomarker of systemic biological aging, has been implicated in cognitive outcomes, but evidence in stroke populations remains limited. We investigated whether baseline LTL and an LTL polygenic risk score (LTL-PRS) are associated with PSCI risk in a large Chinese stroke cohort. METHODS:We analyzed patients with ischemic stroke or transient ischemic attack (TIA) from the Third China National Stroke Registry-impairment of cognition and sleep (CNSR-III-ICONS) sub study. Baseline LTL was estimated from whole-genome sequencing (WGS) using TelSeq. An 18-SNP LTL-PRS (East-Asian weights) was z-standardized. Cognitive status at 12 months was assessed using the Montreal Cognitive Assessment (MoCA); PSCI was defined as an age-, sex-, education-adjusted norm-referenced threshold (z score ≤ -1.5). Fifty-one participants missing education required for the adjusted PSCI definition were excluded, leaving 923 for analysis. Multivariable logistic regression tested associations of LTL and LTL-PRS with PSCI, with exploratory interaction and stratified analyses by age, sex, and stroke subtype. RESULTS:PSCI occurred in 191 of 923 patients (20.7%). Neither continuous LTL nor LTL-PRS was associated with PSCI overall. Exploratory analyses provided nominal evidence of heterogeneity by age (p-interaction = 0.038) and stroke subtype (p-interaction = 0.01), whereas evidence for sex interaction was inconclusive (p-interaction = 0.07). Associations with the MoCA executive function domain were weak and sensitive to covariate adjustment. CONCLUSIONS:Baseline telomere metrics showed, at most, modest and context-dependent associations with PSCI. Neither LTL nor this limited-variant LTL-PRS demonstrated overall prognostic utility, highlighting the need for integrated, longitudinal multi-omic approaches for individualized cognitive risk prediction after stroke.

To develop and validate a deep learning system (DLS) model predicting hematoma expansion (HE) based on non-contrast (NC) CT and a score combining with clinical variables. The multicenter retrospective dataset (R), the multicenter prospective dataset (P1), and the single-center prospective dataset (P2) enrolled 2350, 460, and 96 intracerebral hemorrhage (ICH) patients for analysis, respectively. The DLS model was developed, validated, and tested in R-development (R-dev), R-validation (R-val), and P1, respectively. After exploring clinical predictors of HE using multivariable logistic regression on P1-development (P1-dev), a five-point score “ARCHES” (Ai-Reinforced intraCerebral Hemorrhage hematoma Expansion Score) combining clinical predictors with the DLS model was created. We compared the discrimination of the ARCHES, DLS, with other models using the receiver operating characteristic (ROC) and DeLong test. The areas under the curve (AUC) of the DLS model were 0.781 (95

BACKGROUND:We aimed to explore whether the benefits of tenecteplase within 4.5 to 24 h would be modified by the infarct growth rate (IGR). METHODS:This study is a secondary analysis of the Tenecteplase Reperfusion Therapy in Acute Ischemic Cerebrovascular Events-III (TRACE-III) trial, a Phase 3, multicenter, prospective, open-label, randomized, blinded-outcome-assessment trial. In the TRACE-III trial, patients with large-vessel occlusion at 4.5 to 24 h without thrombectomy were enrolled at 58 centers in China and randomly assigned to receive 0.25 mg/kg tenecteplase or standard medical treatment. Of these, patients with witnessed stroke onset time were analyzed. The IGR was calculated as baseline ischemic core volume divided by time since onset. The primary outcome was the modified Rankin Scale (mRS) 0-1 at 90 days. The treatment effect of tenecteplase versus standard medical treatment was assessed in two groups based on the median IGR cut point. A multiplicative interaction term for IGR*treatment was used to test for effect modification. RESULTS:In 292 eligible patients, the median IGR was 1.4 ml/h. A total of 146 patients with IGR < 1.4 ml/h were classified as ultraslow IGR (Tenecteplase, 70; Standard medical treatment, 76), and 146 patients with IGR ⩾ 1.4 ml/h were classified as slow IGR (Tenecteplase, 73; Standard medical treatment, 73). The rate of no disability (mRS ⩽ 1) was significantly increased with tenecteplase in slow progressors (34.2% vs 15.1%; OR = 2.94, 95% CI 1.32 to 6.55; P = 0.009). The functional status was similar in ultraslow progressors (mRS 0-1: 37.1% vs 35.5%; OR = 1.07, 95% CI 0.55 to 2.11; P = 0.84). A P value for the interaction between IGR and treatment was 0.06. The incidence of sICH and mortality did not differ significantly between the two treatment regimens in the IGR groups. CONCLUSION:Patients with late-window ischemic stroke who are not eligible for endovascular thrombectomy may derive greater benefit from tenecteplase compared with standard medical treatment among those with relatively faster infarct growth, although this finding should be considered exploratory.

Background The role of renal function in patients with acute ischaemic stroke (AIS) receiving blood pressure (BP) lowering therapy remains incompletely understood. In this prespecified subgroup analysis of the CATIS-2 trial, we tested the hypothesis that baseline estimated glomerular filtration rate (eGFR) affects the relationship between early antihypertensive therapy and outcomes in AIS.Methods In the CATIS-2 trial, AIS patients were randomly assigned to early treatment (initiating BP-lowering agents within 24–48 hours of stroke onset) or delayed treatment (withholding such agents until day 8). The composite of functional dependence or death (modified Rankin Scale score ≥3) at day 90 was defined as the primary outcome. The treatment effect was analysed between different baseline eGFR values (<90 versus ≥90 mL/min per 1.73 m²).Results 4685 patients were finally enrolled. Among 1061 patients with eGFR<90, the risk of the primary outcome was higher in early treatment compared with delayed treatment (15.3% vs 10.5%; OR (OR) 1.55, 95% CI 1.07–2.23; p=0.02). In 3624 patients with eGFR≥90, no such difference was observed (10.7% vs 10.2%; OR 1.05, 95% CI 0.85 to 1.30; p=0.63). The interaction was marginally significant (p=0.07).Conclusions The modification effect of renal function on different antihypertensive timing for AIS was not observed. Further, adequately powered studies were warranted.Trial registration number NCT03479554.

Over the past three decades, thrombolytic therapy for acute ischaemic stroke has evolved dramatically since the landmark National Institute of Neurological Disorders and Stroke (NINDS) trial in 1995, which formally established recombinant tissue plasminogen activator as an effective treatment for stroke. This evolution has occurred along three key dimensions: (1) an expanding repertoire of thrombolytic agents, (2) a progressive broadening of the therapeutic window and (3) organisational and technological innovations aimed at minimizing prehospital and in-hospital treatment delays.In this review, we summarise the major milestones in the clinical evidence supporting thrombolytic therapy over the past 30 years, with particular emphasis on large phase III randomised clinical trials. Our goal is to delineate the trajectory of progress in stroke thrombolysis and to provide clinicians and researchers with a clear and coherent framework for understanding the field’s past achievements and future directions.

BACKGROUND:Optimal alteplase dosing for intravenous thrombolysis in stroke patients aged >80 years remains debated due to underrepresentation in pivotal trials. This study evaluates the relative effectiveness and safety of low-dose (0.6 mg/kg) versus standard-dose (0.9 mg/kg) alteplase in elderly patients treated within 4.5 hours of onset. METHODS:Utilizing the China Stroke Center Alliance registry database (2018-2022), we conducted a propensity score-matched analysis encompassing 6,522 patients over 80 years of age, ultimately generating 2,050 matched pairs. The primary outcome was functional dependence or death at discharge (mRS 2-6). Secondary measures included major disability (mRS 3-6), in-hospital mortality, and symptomatic intracranial hemorrhage (sICH). RESULTS:The primary outcome showed no statistical difference between low-dose (81.6 %) and standard-dose (81.8 %) groups (OR 0.99; 95 % CI 0.84-1.16; P=0.871). Rates of major disability (56.0 % vs. 57.7 %; OR 1.02; P=0.730), mortality (15.6 % vs. 14.4 %; OR 1.10; P=0.294), and sICH (4.4 % vs. 5.2 %; OR 0.85; P=0.274) were broadly alike. Low-dose alteplase showed a numerical decline in bleeding events. Subgroup analysis found no modifying factors. CONCLUSIONS:In octogenarian stroke patients, low-dose alteplase offers parallel efficacy and safety to standard therapy, with a possible reduction in hemorrhagic complications. These results support low-dose regimens as a reasonable option for elderly patients, particularly when minimizing bleeding risk is a clinical priority.

BACKGROUND:Evidence links eosinopenia to poor 3-month outcomes after stroke, but its long-term prognostic value in acute ischemic cerebrovascular events remains unclear. This study aimed to evaluate the association between eosinophil counts (and ratios) and 5-year clinical outcomes in patients with acute ischemic stroke or transient ischemic attack (TIA). METHODS:We analyzed data from the Third China National Stroke Registry (CNSR-III). Patients with acute ischemic stroke or TIA were categorized into quartiles based on eosinophil counts and ratios measured within 24 h of admission. Using the highest quartile as the reference, we calculated hazard ratios (HRs) or odds ratios (ORs) with 95% confidence intervals (CIs) for adverse outcomes. The association of eosinophil levels with 5-year risks of stroke recurrence, poor functional outcome, all-cause death, and composite vascular events was evaluated. RESULTS:A total of 12,745 participants were enrolled. Compared to the fourth quartile of eosinophil counts, the first quartile were associated with an increased risk of stroke recurrence (adjusted hazard ratio [HR] = 1.25, 95% confidence interval [CI]: 1.10-1.41) and ischemic stroke recurrence (adjusted HR, 1.25; 95% CI: 1.10-1.42). Similar associations were observed over 5 years for poor functional outcome (adjusted odds ratio [OR] = 1.57, 95% CI: 1.40-1.76), composite vascular events (adjusted HR = 1.15, 95% CI: 1.02-1.29), and all-cause death (adjusted HR = 1.53, 95% CI: 1.32-1.77). Parallel results were found for eosinophil ratios. CONCLUSIONS:This study demonstrated that low levels of both eosinophil counts and eosinophil ratios were associated with an increased risk of adverse clinical outcomes at 5 years of follow-up in patients with ischemic stroke or TIA.

We present a reproducible workflow for real-time visualization of nicotinamide adenine dinucleotide (NAD+) biosensor-channel responses in brain microvascular endothelial cells in living mice using cranial-window two-photon microscopy and an endothelial-targeted fluorescent NAD+ sensor. The protocol includes (1) adeno-associated virus (AAV)-mediated expression of the NAD+ sensor selectively in cerebrovascular endothelium using AAV-X1.1 under the Cdh5 promoter (vascular endothelial cadherin), (2) surgical preparation of a stable 3 mm × 3 mm cortical cranial window, and (3) dual-wavelength two-photon imaging to simultaneously capture the 920 nm-excited green sensor-channel signal and the 1040 nm-excited intravascular tetramethylrhodamine (TMR)-dextran reference channel. The TMR-dextran channel provides a vascular lumen reference and supports vessel selection, motion assessment, and vascular-integrity evaluation. As an application example, we describe nicotinamide mononucleotide-associated changes in endothelial sensor-channel fluorescence following drinking-water, oral-gavage, or intravenous delivery. This protocol emphasizes critical steps, recommended viral-dosing and titer considerations, troubleshooting, and quantitative-analysis strategies, enabling laboratories to implement in vivo monitoring of cerebrovascular NAD+ biosensor-channel dynamics for studies of neurometabolism in health and disease.

Parkinson disease (PD) is an increasing cause of mortality among older adults, but national studies have generally examined mortality trends, forecasts, spatial variation, or environmental associations separately. We integrated these components to characterize U.S. PD mortality and its regional climate correlates. This ecological time-series study used CDC WONDER mortality data for U.S. adults aged 65 years or older from 1999 to 2024. Joinpoint regression, exponential smoothing state-space forecasting with rolling-origin validation, state-level Moran's I, lagged count models, nonlinear spline analyses, and principal component analysis of regional climate metrics were applied. Nominal and Benjamini-Hochberg-adjusted P values were reported. From 1999 to 2024, PD deaths increased from 14,225 to 39,671, and the age-adjusted mortality rate increased from 41.42 to 71.58 per 100,000 population (average annual percent change, 1.99

AIMS:To evaluate the relationship between metabolic syndrome and the efficacy and safety of genotype-guided dual antiplatelet therapy (DAPT) in the CHANCE-2 trial. MATERIALS AND METHODS:This post hoc study used data from the CHANCE-2 trial. Patients with minor stroke or TIA who carried the CYP2C19 loss-of-function (LOF) allele were randomized to receive ticagrelor-aspirin (TIC-ASA) or clopidogrel-aspirin (CLO-ASA). The primary efficacy outcome was stroke recurrence within 90 days. The primary safety outcome was severe or moderate bleeding within 90 days. We classified patients into metabolic syndrome (MetS) and non-metabolic syndrome (non-MetS) groups. Differences in the outcome during 90-days follow-up period were assessed using the Cox proportional hazards model. RESULTS:Among 5652 patients, 3305 were non-MetS and 2347 were MetS. Compared with CLO-ASA, TIC-ASA significantly reduced the risk of stroke recurrence within 90 days among patients with MetS (6.44% vs. 9.90%; HR 0.64, 95% CI 0.48-0.85; p < 0.01); this benefit was not seen in non-MetS (6.03% vs. 5.96%; HR 1.01, 95% CI 0.77-1.34; p = 0.93), with a significant interaction effect (p for interaction = 0.03). Moreover, a linear trend was observed (p for trend = 0.04), indicating metabolic health status may modify the efficacy of genotype-guided DAPT in a dose-dependent manner. No significant difference was observed in severe or moderate bleeding events by metabolic syndrome (MetS: 0.42% vs. 0.35%, non-MetS: 0.25% vs. 0.36%; p for interaction = 0.55). CONCLUSIONS:Among CYP2C19 LOF carriers with minor stroke or TIA, patients with MetS received more clinical benefit from TIC-ASA versus CLO-ASA compared to those with non-MetS. Registry: ClinicalTrials.gov, TRN: NCT04078737.

BACKGROUND:Atherosclerosis is a major factor in cognitive decline among aging individuals and is frequently linked to the accumulation of white matter hyperintensities. Brain resilience, which represents the brain's capacity to withstand external disruptions, remains poorly understood in terms of how atherosclerosis impacts it and, in turn, influences cognition. Here, we investigated the relationship between atherosclerosis, white matter hyperintensities, and structural network resilience, along with their combined effects on cognitive performance. METHODS:We utilized data from the large-scale community cohort Polyvascular Evaluation for Cognitive Impairment and Vascular Events (n = 2160). Whole-brain structural connections were constructed, and structural disconnections were simulated based on white matter hyperintensities. SNR, serving as a marker to quantify structural network resilience, is defined by the similarity of hub nodes between the original network and its disconnected counterpart. RESULTS:SNR showed higher odds ratios compared to white matter hyperintensities in relation to arterial status. Additionally, chain mediation analysis indicated that cognitive decline associated with atherosclerosis was partially mediated by both white matter hyperintensities and structural network resilience. Atherosclerosis accelerates the degradation of brain structural network resilience as age increases. CONCLUSIONS:These findings suggest that SNR could offer complementary insights into cognitive decline caused by atherosclerosis and serve as a potential biomarker of brain health in atherosclerotic conditions. Additionally, SNR may act as an indicator for guiding the selection of future therapies for atherosclerosis.

Background The incidence of vascular cognitive impairment (VCI) is high in patients suffering from ischaemic stroke or transient ischaemic attack (TIA) or with vascular risk factors. Effective prevention strategies for VCI remain limited. Anaemia or low haemoglobin was found as an independent risk factor for adverse outcomes after acute stroke. Anaemia or low haemoglobin was possibly associated with an increased risk of poststroke cognitive impairment. Whether supplement of ferrous iron to correct anaemia reduces the risk of VCI and improves adverse outcomes in patients with ischaemic cerebrovascular disease remains uncertain.Aim We aim to introduce the design and rationale of the safety and efficacy of Ferrous iron on the prevention of Vascular cOgnitive impaiRment in patients with cerebral Infarction or TIA (FAVORITE) trial.Design FAVORITE is a randomised, placebo-controlled, double-blind, multicentre trial that compares supplement of ferrous iron with placebo for recent minor stroke/TIA patients complicated with mild anaemia or iron deficiency: Ferrous succinate sustained-release tablet 0.2 g (corresponding to 70 mg of elemental iron) once daily after or during breakfast for 12 weeks or placebo with much the same colour, smell and size as ferrous iron once daily during or after breakfast for 12 weeks. All paticipants will be followed within the next year.Study outcomes The primary effective outcome is the incidence of VCI at 3 months after randomisation and the primary safety outcome includes any gastrointestinal adverse event during 3 months.Discussion The FAVORITE trial will clarify whether supplement of ferrous iron to correct low haemoglobin reduces the risk of VCI in patients with recent ischaemic stroke or TIA complicated with mild anaemia or iron deficiency compared with placebo.Trial registration number NCT03891277

Background:The effect of colchicine on subsequent stroke among patients with and without symptomatic intracranial artery stenosis (sICAS) and whether age modifies such effect are not known. Methods:In this prespecified subgroup analysis of CHANCE-3, a randomized, double-blind, placebo-controlled clinical trial conducted at 244 centers in China between 11 August 2022 and 13 April 2023 (ClinicalTrials.gov number, NCT05439356), we included 7567 patients with ischemic stroke or transient ischemic attack and assessments of intracranial arteries at baseline. The primary efficacy outcome was a new stroke at 90 days. The main secondary outcome was a combined vascular event including ischemic stroke, hemorrhagic stroke, TIA, myocardial infarction, and vascular death. The primary safety outcome was any serious adverse event within 90 days. Findings:In patients with sICAS, 141 (10.5%) patients on colchicine and 115 (8.5%) on placebo had recurrent stroke within 90 days (adjusted HR 1.30, 95% CI 1.01-1.69; p = 0.04); in patients without sICAS, the corresponding event rates were 4.2% and 5.2% (adjusted HR 0.80, 95% CI 0.62-1.05; p = 0.10) (adjusted interaction p = 0.01). A significant interaction was also observed between sICAS status and the effect of colchicine on the secondary outcome of combined vascular events (adjusted p = 0.02). The interaction was more apparent in the elderly patients (adjusted p < 0.001). In the elderly patients with sICAS (n = 1648), the risk of stroke was higher in the colchicine group (n = 829, 50.3%) compared to the placebo group (n = 819, 49.7%) (adjusted HR 1.58, 95% CI 1.13-2.20; p < 0.001). There was no interaction of status of sICAS with treatment groups on primary safety outcome of any serious adverse event (p = 0.54). In patients with sICAS, 24 (1.8%) patients on colchicine and 7 (0.5%) on placebo had diarrhea within 90 days (p = 0.002). In patients without sICAS, diarrhea occurred in 43 (1.8%) patients on colchicine and in 20 (0.8%) patients on placebo (p = 0.004). Interpretation:The effect of colchicine on subsequent stroke within 90 days may differ according to the presence of sICAS. Aging might be associated with an increased risk of early recurrent stroke in the patients with sICAS receiving colchicine treatment. Future prospective studies are needed to confirm these results. Funding:National Key R&D Program of China, National Natural Science Foundation of China, the Capital's Funds for Health Improvement and Research and Chinese Academy of Medical Sciences Innovation Fund for Medical Sciences. China Kunming Pharmaceuticals supplied colchicine and placebo. Guangdong Wesail Biotech Co. provided assistance in measurement of hsCRP levels.

BACKGROUND:A comprehensive epidemiological investigation of the coexistence between diabetes and stroke/TIA in China is urged. METHODS:Data from the Chinese Stroke Center Alliance program, a nationwide multi-center registry study, were used to detect the prevalence, awareness, treatment, and control of diabetes among stroke/TIA. The distribution of diagnosed and undiagnosed diabetes and prediabetes among stroke/TIA patients was investigated, the medical care around diabetes and their respective risk predictors were analyzed, and the association of all above diabetes characteristics with in-hospital death was evaluated using multi-variable Cox regression models. RESULTS:Of 980 625 patients included, 308 426 (31.5%) had prediabetes, while 365 052 (37.2%) had diabetes, with nearly a third of them undiagnosed (112 969, 30.9%). Of residual aware diabetic patients, 59.0% were treated, with 27.3% controlled. Compared to Han ethnicity, Zhuang ethnicity had a lower prevalence of diabetes (37.3% vs. 35.1%) but were less aware (69.4% vs. 56.5%), treated (59.4% vs. 47.8%), and controlled (27.4% vs. 26.0%). Patients with prediabetes, diagnosed, and undiagnosed diabetes had increasingly higher risks of in-hospital death (adjusted HR [95% CI]: 1.47 [1.35-1.60]; 2.15 [1.97-2.34]; 4.20 [3.87-4.56], all p < 0.001). Unaware and untreated diabetes were independently associated with in-hospital death (adjusted HR [95% CI]: 1.99 [1.85-2.14]; 2.84 [2.63-3.07, both p < 0.001]). Compared with controlled diabetes, those with uncontrolled diabetes had a lower risk of in-hospital death (adjusted HR [95% CI]: 0.77[0.68-0.88], p < 0.001). CONCLUSIONS:The findings indicate that over two-thirds of stroke/TIA patients are exposed to diabetes in China, causing higher in-hospital mortality, which should be screened and intervened early.

Background This study aimed to investigate the relationship between overall obesity, central obesity and brain volumes, as well as to determine the extent to which cardiometabolic and inflammatory measures act as mediators in the association between body mass index (BMI), waist-hip ratio (WHR) and brain volumes.Methods In the context of counterfactual framework, mediation analysis was used to explore the potential mediation in which cardiometabolic and inflammatory measures may mediate the relationship between BMI, WHR, and brain volumes.Results Among 2413 community-dwelling participants, those with high BMI or WHR levels experienced an approximately brain ageing of 4 years. Especially, individuals with high WHR or BMI under the age of 65 exhibited white matter hyperintensity volume (WMHV) differences equivalent to around 5 years of ageing. Conversely, in the high-level WHR population over the age of 65, premature brain ageing in gray matter volume (GMV) exceeded 4.5 years. For GMV, more than 45% of the observed effect of WHR was mediated by glycaemic metabolism indicators. This proportion increases to 78.70% when blood pressure, triglyceride, leucocyte count, and neutrophil count are jointly considered with glycaemic metabolism indicators. Regarding WHR and BMI’s association with WMHV, cardiometabolic and inflammatory indicators, along with high-density lipoprotein cholesterol, mediated 35.50% and 20.20% of the respective effects.Conclusions Overall obesity and central obesity were associated with lower GMV and higher WMHV, a process that is partially mediated by the presence of cardiometabolic and inflammatory measures.

BackgroundThe COVID-19 pandemic has highlighted the crucial role of artificial intelligence (AI) in predicting mortality and guiding healthcare decisions. However, AI models may perpetuate or exacerbate existing health disparities due to demographic biases, particularly affecting racial and ethnic minorities. The objective of this study is to investigate the demographic biases in AI models predicting COVID-19 mortality and to assess the effectiveness of transfer learning in improving model fairness across diverse demographic groups.MethodsThis retrospective cohort study used a population-based dataset of COVID-19 cases from the Centers for Disease Control and Prevention (CDC), spanning the years 2020-2024. The study analyzed AI model performance across different racial and ethnic groups and employed transfer learning techniques to improve model fairness by adapting pre-trained models to the specific demographic and clinical characteristics of the population.ResultsDecision Tree (DT) and Random Forest (RF) models consistently showed improvements in accuracy, precision, and ROC-AUC scores for Non-Hispanic Black, Hispanic/Latino, and Asian populations. The most significant precision improvement was observed in the DT model for Hispanic/Latino individuals, which increased from 0.3805 to 0.5265. The precision for Asians or Pacific Islanders in the DT model increased from 0.4727 to 0.6071, and for Non-Hispanic Blacks, it rose from 0.5492 to 0.6657. Gradient Boosting Machines (GBM) produced mixed results, showing accuracy and precision improvements for Non-Hispanic Black and Asian groups, but declines for the Hispanic/Latino and American Indian groups, with the most significant decline in precision, which dropped from 0.4612 to 0.2406 in the American Indian group. Logistic Regression (LR) demonstrated minimal changes across all metrics and groups. For the Non-Hispanic American Indian group, most models showed limited benefits, with several performance metrics either remaining stable or declining.ConclusionsThis study demonstrates the potential of AI in predicting COVID-19 mortality while also underscoring the critical need to address demographic biases. The application of transfer learning significantly improved the predictive performance of models across various racial and ethnic groups, suggesting these techniques are effective in mitigating biases and promoting fairness in AI models.