Category: Statistics

Ultrason Imaging. 2026 Aug 19:1617346261476280. doi: 10.1177/01617346261476280. Online ahead of print.

ABSTRACT

To compare pykan and Lightweight KAN with MLP and Linear classifier heads for focal liver lesion classification on B-mode ultrasound, with emphasis on diagnostic sensitivity, interpretability and computational cost. A frozen ResNet18 feature extractor was combined with 4 classifier heads and evaluated on 735 annotated liver ultrasound images (Benign, Malignant, Normal) using 5-fold stratified cross-validation. Weighted cross-entropy was applied to address class imbalance. Performance was assessed using Accuracy, F1-Macro, F1-Weighted, AUROC, class-wise recall, malignant-to-normal misclassification count, inference latency, and pykan-specific spline activeness analysis. Overall metrics were similar across models, and Wilcoxon Signed-Rank testing showed no statistically significant pairwise differences (p > .05). However, clinically relevant class-level differences emerged. pykan achieved the highest Benign recall (65.5%) and the lowest malignant-to-normal error rate (5/435, 1.1%). Lightweight KAN achieved the highest Malignant recall (82.1%) and markedly faster inference (0.69 ms/image) than pykan (4183.67 ms/image). Aggregate metrics alone do not capture clinically meaningful trade-offs among KAN implementations. Lightweight KAN is attractive for rapid screening, whereas pykan shows a directional trend toward fewer malignant-to-normal errors and a more interpretable profile, though this trend did not reach statistical significance and requires confirmation in larger studies.

PMID:42618538 | DOI:10.1177/01617346261476280

Technol Health Care. 2026 Aug 19:9287329261479267. doi: 10.1177/09287329261479267. Online ahead of print.

ABSTRACT

Modern healthcare system produces large amount of sensitive clinical data, but no existing framework simultaneously addresses and satisfy rigorous security guarantees, storage efficiency and the throughput demand of clinical operations/transactions. All existing framework addresses these requirements in isolation such as cloud architectures lack cryptographic immutability and standalone blockchain systems suffer by high gas costs and block size limitations. To resolve this gap, we present the Blockchain-Integrated Cloud Compression Model (BICCM) a three-tier architecture that combines (i) Ethereum based smart contracts with Role-Based Access Control (RBAC) and Zero-Knowledge Proof (ZKP) authentication, (ii) Lempel-Ziv-Welch (LZW) lossless compression applied at a dedicated light node layer and (iii) a blockchain secured cloud object store with on-chain SHA-256 has anchoring. COVID-19 medical records from Pakistan Bureau of statistics (PBS-2023) in six size classes range from 4.93 to 39.80KB were used for evaluation.BICCM achieves 65.81 to 85.34% storage reduction, 64.3 to 84.4% gas savings (mean 76.4%), 45.4 to 83.2% execution-time improvement, and 50% transaction cost reduction Compared to the traditional Ethereum blockchain outperformed four competing algorithms (LZ-String, LZ77, Run-Length Encoding, Huffman) for all evaluated data sizes, attaining compression ratios of 2.92 to 6.82. Smart contract deployment on auxiliary nodes reduced ETH costs by 88.6% compared to the main deployment node. BICCM offers a scalable, safe and HIPAA/GDPR complaint platform for blockchain cloud healthcare data management with full implementation across all twelve assessed security criteria. Future work will follow a strategy for federated learning integration and post-quantum cryptography transition.

PMID:42618521 | DOI:10.1177/09287329261479267

J Neurosci. 2026 Aug 19:e0325262026. doi: 10.1523/JNEUROSCI.0325-26.2026. Online ahead of print.

ABSTRACT

Human vision efficiently navigates information-dense environments by extracting summary statistics-the average properties of item groups-to circumvent capacity limits. However, the neural transition from location-specific sensory registration to location-invariant abstract representation remains poorly understood. We recorded high-density EEG while participants of either sex performed an ensemble size discrimination task, using time-resolved multivariate pattern analysis (MVPA) and cross-visual-field generalization to dissociate these two levels of representation. Our results reveal a clear temporal hierarchy: location-specific ensemble size information emerged as early as ∼40 ms post-stimulus, significantly preceding the onset of location-generalized, abstract representation at ∼89 ms. During the early location-specific decoding phase, we observed a distinct left-visual-field advantage, with higher neural decoding accuracy predicting superior behavioral precision. Critically, error trials were characterized by premature neural generalization, suggesting an inherent trade-off: while abstraction is essential for efficient summarization, sacrificing sensory fidelity too early impairs perceptual accuracy. Furthermore, distinct oscillatory mechanisms supported this transformation-low-frequency (delta/theta) activity underpinned location-specific encoding, whereas mid-frequency (alpha/beta) oscillations robustly sustained location-invariant abstraction. Individual differences in the strength of these late-stage location-specific and location-invariant representations were predictable from intrinsic resting-state occipito-parietal gamma power. Together, these findings provide a comprehensive neural model of ensemble perception, demonstrating that the brain constructs abstract summary statistics through a temporally ordered, spectrally specific transformation that balances sensory fidelity with representational abstractionSignificance Statement How the brain transforms complex sensory input into simplified summary statistics-like the average size of a group-is fundamental to efficient vision. This study reveals a critical temporal and spectral hierarchy in this process. Using high-density EEG and multivariate analysis, we show that the brain first registers ensemble information at specific locations before transforming it into a location-invariant, abstract representation. We demonstrate that premature abstraction leads to perceptual errors, suggesting a vital balance between sensory detail and abstract summary. Furthermore, we identify distinct neural oscillations that coordinate this transformation. These findings provide a new model for how the human brain constructs stable, abstract representations from a dynamic and cluttered visual world.

PMID:42618511 | DOI:10.1523/JNEUROSCI.0325-26.2026

Zhonghua Yi Xue Za Zhi. 2026 Aug 25;106(31):3285-3290. doi: 10.3760/cma.j.cn112137-20250913-02384.

ABSTRACT

A retrospective analysis was conducted on the neuroimaging features of white matter lesions related to patent foramen ovale and clinical characteristics (age, sex, history of hypertension and diabetes, etc.) of patients who underwent patent foramen ovale closure surgery at Fuwai Hospital, Chinese Academy of Medical Sciences, from February 2021 to June 2025. All patients were confirmed to have right-to-left shunt by contrast-enhanced transthoracic echocardiography and completed brain MRI examination. According to the results of contrast-enhanced transthoracic echocardiography, patients were divided into 2 groups (large-shunt group and non-large-shunt group). The distribution characteristics of white matter lesions were investigated from multiple dimensions, including anterior/posterior circulation, left/right hemisphere, and single/multiple lesions. A total of 118 patients were screened, among whom 100 with complete clinical data were included in the study. The age was (45.4±9.9) years. There were 62(62%) cases in the large-shunt group and 38(38%) cases in the non-large-shunt group. No statistically significant differences were found between the 2 groups in terms of age, gender, or history of hypertension, diabetes, stroke or transient ischemic attack (TIA). The proportion of multiple white matter lesions in the large-shunt group was significantly different [59.7%(37/62) vs 39.5%(15/38), P=0.049], as was the proportion of bilateral hemispheric distribution [62.9%(39/62) vs 42.1%(16/38), P=0.042]. The proportion of lesions distributed in the right hemisphere was higher [66.1%(41/62) vs 47.4%(18/38)], but the difference did not reach statistical significance (P=0.064). Among patients with a history of stroke or TIA, the proportions of multiple, bilateral, and involvement of both anterior and posterior circulation white matter lesions were significantly higher than those without such histor[66.1%(39/59) vs 31.7%(13/41), 67.8%(40/59) vs 36.6%(15/41), 62.7%(37/59) vs 26.8%(11/41), all P<0.01]. No statistically significant difference in white matter lesions distribution was observed between patients with and without a history of migraine. Patients with patent foramen ovale and a large right-to-left shunt are more likely to present with bilateral and multiple cerebral white matter lesions on brain MRI, and these white matter lesions are more prone to occur in the right hemisphere.

PMID:42618505 | DOI:10.3760/cma.j.cn112137-20250913-02384

Zhonghua Yi Xue Za Zhi. 2026 Aug 25;106(31):3273-3284. doi: 10.3760/cma.j.cn112137-20251205-03203.

ABSTRACT

Objective: To investigate the effect and mechanism of chloroquine on the growth of pancreatic neuroendocrine tumors (pNET). Methods: This study was divided into animal experiments and cell experiments. (1) Animal experiments: ten male BALB/c nude mice, 6-8 weeks old, were selected. The suspension of QGP-1 cells (human pNET cell line) in the logarithmic growth phase was inoculated subcutaneously into the right axilla of nude mice to establish a subcutaneous xenograft tumor model. According to the random number table method, the mice were divided into a control group (intraperitoneal injection of physiological saline, once a day) and a chloroquine group (intraperitoneal injection of 50 mg/kg chloroquine, once a day), with 5 mice in each group. The treatment was administered continuously for 4 weeks. At the end of the 4th week, the mice were sacrificed, and the long diameter and weight of the transplanted tumors in each group were compared. The expression level of Ki-67 in tumor tissues was detected by immunohistochemistry (IHC) method, and tumor cell apoptosis was detected by TUNEL fluorescence staining. (2) Cell experiments: the cells used were human pNET cell line QGP-1 and pNET tissues (primary PNET cells) isolated from patients diagnosed with pNET at the First Affiliated Hospital of Nanjing Medical University who underwent surgical resection. Primary PNET cells were treated with 0, 10, 20, 40, 80, and 100 μmol/L chloroquine, and were respectively named the blank group [adding an equal volume of dimethyl sulfoxide (DMSO)] and 10, 20, 40, 60, 80, 100 μmol/L groups; QGP-1 cells were treated with 0, 20, 40, 60, 80, 100, and 160 μmol/L chloroquine, and were respectively named the blank group (adding an equal volume of DMSO) and 20, 40, 60, 80, 100, 160 μmol/L groups. After 24 hours of treatment, cell viability (absorbance A₄₅₀ nm value) was detected by cell counting kit-8 (CCK-8) method, and the half-maximal inhibitory concentration (IC50) was calculated. Based on the IC50 results, and combined with the long-term non-specific cytotoxicity, low-toxicity sub-lethal concentrations were selected for intervention in subsequent functional experiments. Cell proliferation ability was assessed by 5-ethynyl-2′-deoxyuridine (EdU) incorporation assay (proportion of EdU-positive cells) and plate colony formation assay (number of cloned cell formations). Cell migration (number of migrating cells) and invasion ability (number of invading cells) were detected using Transwell method. Cell cycle (proportion of cells in G0 or G1 phase) and apoptosis (early and late cell apoptosis rates) were analyzed by flow cytometry. Differentially expressed genes were screened using transcriptome sequencing, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis was performed. The expression levels of pathway-related phosphorylated proteins [phosphorylated phosphatidylinositol 3-kinase (p-PI3K), phosphorylated protein kinase B (p-AKT), phosphorylated mammalian target of rapamycin (p-mTOR)] and autophagy-related proteins [ubiquitin-binding protein 62 (P62), key autophagy molecule Beclin-1 (Beclin1-1) and microtubule-associated protein 1 light chain 3-Ⅱ (LC3-Ⅱ)] were detected by Western blotting (WB). Results: (1) Animal experiments: tumors grew in both the control group and the chloroquine group, indicating successful modeling. The long diameter of the transplanted tumors in the chloroquine group was lower than that in the control group (P<0.05). There was no statistically significant difference in body weight between the chloroquine group and the control group at different administration times (days 1, 3, 5, 7, and 9 after administration) (all P>0.05). The proportion of Ki-67 positive cells in the chloroquine group was lower than that in the control group, while the proportion of TUNEL-positive cells (the number of apoptotic cells) was higher than that in the control group (P<0.05). (2) Cell experiments: after primary PNET cells were stimulated with different concentrations of chloroquine for 24, 48, and 72 h, the cell viability A₄₅₀ nm values of the 10, 20, 40, 60, 80, and 100 μmol/L groups at each time point were all lower than those of the blank group (all P<0.05); After QGP-1 cells were stimulated with different concentrations of chloroquine for 24, 48, and 72 h, the A₄₅₀ nm values of the 20, 40, 60, 80, 100, and 160 μmol/L groups at each time point were all lower than those of the blank group (all P<0.05). At 24 h of treatment, the IC50 values of primary PNET cells and QGP-1 cells were 67 and 101 μmol/L, respectively; therefore, in subsequent experiments, the blank group, 10 μmol/L group, and 20 μmol/L group were selected for primary PNET cells, and the blank group, 20 μmol/L group, and 40 μmol/L group were selected for QGP-1 cells. The proportion of EdU-positive cells, the number of cloned cell formations, and the number of migrating and invading cells in the 10 μmol/L and 20 μmol/L primary PNET cell groups were all less than those in the blank group, while the proportion of cells in the G1 phase and the apoptosis rate were all higher than those in the blank group (all P<0.05); additionally, the proportion of EdU-positive cells, the number of cloned cell formations, and the number of migrating and invading cells in the 20 μmol/L group of primary PNET cell were all less than those in the 10 μmol/L group, while the proportion of cells in the G0 or G1 phase and the (early and late) cell apoptosis rates were all higher than those in the 10 μmol/L group (all P<0.05). The proportion of EdU-positive proliferating cells, the number of cloned cell formations, and the number of migrating and invading cells in the 20 μmol/L and 40 μmol/L group of QGP-1 cell were all less than those in the blank group, while the proportion of cells in the G0 or G1 phase and the (early and late) cell apoptosis rates were all higher than those in the blank group (all P<0.05); additionally, the proportion of EdU-positive proliferating cells, the number of cloned formations, and the number of migrating and invading cells in the 40 μmol/L group of QGP-1 cell were all less than those in the 20 μmol/L group, while the proportion of cells in the G0 or G1 phase and the (early and late) cell apoptosis rates were all higher than those in the 20 μmol/L group (all P<0.05). Transcriptome sequencing analysis showed that differentially expressed genes were enriched in the phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT)/mammalian target of rapamycin (mTOR) pathway; WB method results showed that the relative expression levels of phosphorylated proteins such as p-mTOR, p-PI3K, and p-AKT in the 20 μmol/L group of QGP-1 cells were lower than those in the blank group (all P<0.05); the relative expression levels of the autophagy substrate P62 and the autophagosome marker LC3-Ⅱ protein were higher than those in the blank group, while Beclin-1 was lower than that in the blank group (all P<0.05). Conclusion: Chloroquine may inhibit the occurrence and development of pNET by suppressing the phosphorylation activation of the PI3K/AKT/mTOR signaling pathway and blocking autophagy.

PMID:42618504 | DOI:10.3760/cma.j.cn112137-20251205-03203

Zhonghua Yi Xue Za Zhi. 2026 Aug 25;106(31):3256-3263. doi: 10.3760/cma.j.cn112137-20251230-03481.

ABSTRACT

Objective: To compare the screening efficacy of different evaluation indicators of peak expiratory flow (PEF) for chronic obstructive pulmonary disease (COPD), and to assess the consistency of different evaluation indicators of PEF with forced expiratory volume in one second (FEV₁) measured value% predicted value when grading the severity of airflow limitation. Methods: A retrospective analysis was conducted on data from 1 512 patients clinically suspected of COPD who underwent pulmonary function test at the Pulmonary Function Laboratory of China-Japan Friendship Hospital between May 2023 and May 2024. Pulmonary function test results were collected, including post-bronchodilator FEV₁/forced vital capacity (FVC), pre-bronchodilator FEV₁ measured value% predicted value, pre-bronchodilator PEF measured value, pre-bronchodilator PEF measured value divided by height², pre-bronchodilator PEF measured value % predicted value, post-bronchodilator PEF measured value, post-bronchodilator PEF measured value/height², and post-bronchodilator PEF measured value% predicted value. Receiver operating characteristic (ROC) curve was used to evaluate the screening performance of different PEF indicators for detecting impaired lung function, with calculation of the area under the curve (AUC), differences in AUC among indicators, and 95%CI for both AUC values and their differences. Kappa consistency tests were performed to assess agreement between PEF measured value/height² and PEF measured value% predicted value with FEV₁ measured value% predicted value in evaluating the severity of airflow limitation. Results: Among 1 512 clinically suspected patients with COPD, 1 258 were male and 254 were female, with age of (65.68±9.02) years. The AUC (95%CI) for different PEF indicators in screening COPD, ranked from highest to lowest, were as follows: pre-medication PEF measured value% predicted value [0.881 (0.861-0.901)], post-medication PEF measured value% predicted value [0.879 (0.858-0.899)], pre-medication PEF measured value/height² [0.869 (0.847-0.890)], post-medication PEF measured value/height² [0.865 (0.843-0.887)], pre-medication PEF measured value [0.843 (0.818-0.867)], and post-medication PEF measured value [0.837 (0.813-0.862)]. After pairwise comparison of the six indicators, it was found that regardless of pre-medication or post-medication, the AUC of PEF measured value was the lowest, followed by PEF measured value/height², and the AUC of PEF measured value% predicted value was the highest (except for pre-medication PEF measured value/height² vs post-medication PEF measured value% predicted value: P=0.067, all others P<0.05); and there was no statistically significant difference in the AUC of PEF measured value, PEF measured value/height², and PEF measured value% predicted value before and after medication (all P>0.05). The predictive performance of different PEF indicators for airflow limitation before medication all showed that in extremely severe, mild, and severe airflow limitation, the AUC of PEF measured value% predicted value was the highest, followed by PEF measured value/height², and the AUC of PEF measured value was the lowest (all P<0.05). However, in moderate airflow limitation, there was no significant difference in the predictive performance of different PEF indicators before medication (all P>0.05). The consistency of PEF measured value% predicted value and PEF measured value/height² in evaluating mild airflow limitation was relatively high (Kappa values were 0.671 and 0.611, respectively), and PEF measured value% predicted value was higher than PEF measured value/height²; in moderate, severe, and extremely severe airflow limitation, the consistency of the two indicators with FEV₁ measured value% predicted value was moderate or poor. Conclusions: The percentage of the measured PEF value relative to the predicted value demonstrates the highest screening efficacy for COPD and no bronchodilation test is required; with the exception of mild airflow limitation, PEF is not an appropriate alternative to FEV₁ for classifying the severity of airflow obstruction.

PMID:42618502 | DOI:10.3760/cma.j.cn112137-20251230-03481

Zhonghua Liu Xing Bing Xue Za Zhi. 2026 Aug 10;47(8):1520-1526. doi: 10.3760/cma.j.cn112338-20251108-00805.

ABSTRACT

Environmental pharmacoepidemiology (EPE) is an emerging interdisciplinary field developed to address the health challenges arising from the complex interactions between environmental factors and drug effects. This paper systematically introduces the definitions, core research domains, methodological framework, and public health significance of EPE. The field’s primary objective is to integrate data on environmental exposures and drug utilization to evaluate how the environment influences the efficacy and safety of pharmaceuticals, as well as the potential impacts of drugs on the environment. By leveraging exposomics, phenome-wide data, and real-world big data, while applying advanced statistical methods like propensity scores and multilevel models, EPE facilitates a deeper investigation into the intricate “environment-drug-health” nexus. While facing challenges such as dynamic exposure assessment, poorly understood interaction mechanisms, and statistical modeling limitations, future advancements in EPE will provide crucial scientific support for achieving precision medicine and advancing the synergistic governance of environmental health.

PMID:42618490 | DOI:10.3760/cma.j.cn112338-20251108-00805

Zhonghua Liu Xing Bing Xue Za Zhi. 2026 Aug 10;47(8):1485-1491. doi: 10.3760/cma.j.cn112338-20260105-00013.

ABSTRACT

Objective: To compare the risk of incident diabetic retinopathy (DR) after initiation of glucagon-like peptide-1 receptor agonists (GLP-1RA) versus dipeptidyl peptidase-4 inhibitors (DPP-4i) among patients with type 2 diabetes mellitus (T2DM). Methods: An active-comparator new-user retrospective cohort study was conducted using the Yinzhou Regional Health Information Platform in Ningbo, China. Patients with T2DM who received a first prescription for a GLP-1RA or a DPP-4i between September 1, 2022 and August 31, 2025 were included. 1∶1 nearest-neighbor propensity score matching was applied to balance baseline covariates including demographic characteristics, diabetes duration, baseline comorbidities, prior medication history, and the most recent pre-index measurements of glycated hemoglobin and estimated glomerular filtration rate. In the matched cohort, multivariable Cox proportional hazards regression models were used to estimate hazard ratios (HRs) with 95%CIs to evaluate the association between the initiation of the two medications and the risk of incident DR. Sensitivity analyses were performed by varying the washout period, the latency period, and the matching ratio. Results: After matching, 1 839 patients were included in each group. The incidence rates of DR were 13.60 per 1 000 person-years among DPP-4i and 9.95 per 1 000 person-years among GLP-1RA. Initiation of GLP-1RA was not associated with a statistically significant difference in DR risk compared with initiation of DPP-4i (P=0.114). Findings were consistent across sensitivity analyses. Conclusion: In this real-world cohort, DPP-4i initiation of GLP-1RA was not associated with an increased risk of incident DR compared with initiation of in patients with T2DM.

PMID:42618485 | DOI:10.3760/cma.j.cn112338-20260105-00013

Zhonghua Liu Xing Bing Xue Za Zhi. 2026 Aug 10;47(8):1464-1470. doi: 10.3760/cma.j.cn112338-20260415-00256.

ABSTRACT

Objective: To analyze the epidemic trends and spatiotemporal clustering characteristics of influenza in Shijiazhuang during the 2023-2026 surveillance years, and to provide evidence-based support for the precision prevention and control of influenza. Methods: Data on influenza cases from March 1, 2023 to March 31, 2026 were obtained from the Infectious Disease Reporting Information Management System of Chinese Information System for Disease Control and Prevention. Descriptive epidemiological methods were used to analyze the epidemic characteristics of influenza. Local indicators of spatial association and the SaTScan Poisson-based spatial scan statistic were used for spatial autocorrelation analysis. Results: During the 2023-2026 surveillance years, the annualized incidence rate of influenza in Shijiazhuang was 644.7 per 100 000 person- years, with rates of 370.0 per 100 000 person-years, 616.1 per 100 000 person-years, and 971.9 per 100 000 person-years in 2023, 2024, and 2025 surveillance years, respectively. Students accounted for the largest proportion of cases (30.9%). Three epidemic peaks were identified, occurring in the winters of 2023, 2024, and 2025, and the number of reported cases across these three winter peaks increased year by year. The 2025 influenza season began approximately 2-3 weeks earlier than in the previous two years. Spatial autocorrelation analysis showed a global Moran’s I of 0.375 for the surveillance period, with clusters concentrated in the five main urban districts. The global Moran’s I values of the second and third epidemic peaks were similar (0.360 and 0.359). Clustering began to stabilize from the second epidemic peak, showing a transmission trend of outward diffusion from the five main urban districts. Conclusions: After the gradual withdrawal of non-pharmaceutical interventions during the 2023-2026 surveillance years, influenza in Shijiazhuang showed a continuously expanding epidemic scale, with the three epidemic peaks increasing in size year by year; students were the most affected population. Spatially, influenza transmission showed a pattern of spreading from the five main urban districts to the near suburbs and spilling over to the outer suburbs, for which the transportation hub effect and regular cross-regional commuting may be the main driving factors. It is recommended to strengthen the promotion of influenza vaccination among school-age children before the start of the autumn semester and to enhance school-based influenza surveillance, and to strengthen influenza surveillance in key venues such as high-speed rail stations and airports and among key populations such as regular commuters.

PMID:42618482 | DOI:10.3760/cma.j.cn112338-20260415-00256

Zhonghua Liu Xing Bing Xue Za Zhi. 2026 Aug 10;47(8):1457-1463. doi: 10.3760/cma.j.cn112338-20260211-00119.

ABSTRACT

Objective: To analyze the epidemiological characteristics of Japanese encephalitis (JE) in Henan Province from 2004 to 2025, and the results of antibody among the healthy population and mosquito surveillance at sentinel sites in 2012. Methods: A descriptive epidemiological analysis was conducted of JE case data in Henan Province from 2004 to 2025 obtained from the China Information System for Disease Control and Prevention. In 2012, Luoyang and Xinyang Cities were selected as surveillance sites. Blood samples were collected from healthy individuals across eight age groups, and neutralizing antibodies against the JE virus were detected using a micro-plaque reduction neutralization test. Mosquito specimens were collected using the person-hour method to calculate mosquito density and isolate the JE virus. Results: From 2004 to 2025, the incidence rate of JE in Henan Province decreased from 0.94/100 000 to 0.01/100 000, showing an overall downward trend average annual percentage change (AAPC)=-13.00% (95%CI:-30.00% – -10.00%, P=0.004). The cases mainly occurred from July to September (91.21%, 4 211/4 617). The distribution of cases was predominantly in the southern regions of Henan Province (Nanyang, Xinyang, Luoyang, Zhumadian, Zhoukou, and Shangqiu), accounting for 76.69% (3 541/4 617). Cases aged 0-14 years accounted for 78.32% (3 616/4 617), but the incidence in the population aged ≥15 years showed an increasing trend (AAPC=8.87%, 95%CI: 4.76%-13.71%, P<0.001), with a particularly significant increase in those aged ≥40 years (AAPC=11.93%, 95%CI: 7.32%-16.99%, P<0.001). In Luoyang City, cases aged ≥15 years accounted for 63.62% (376/591), significantly higher than the provincial average (21.68%, 1 001/4 617). The difference in the proportion of cases aged ≥15 years between Henan Province and Luoyang City was statistically significant (χ²=471.69, P<0.001). At surveillance sites in Luoyang City and Xinyang City in 2012, the positive rate of JE antibodies among the healthy population aged ≥15 years in Luoyang City was 53.28% (65/122), significantly lower than that in Xinyang City (91.34%, 116/127), which was consistent with the predominance of cases in the population aged ≥15 years in Luoyang City and cases aged 0-14 years in Xinyang City. The peak mosquito density in Luoyang City occurred in August, which was one month later than in Xinyang City, consistent with the peak JE incidence in Luoyang City occurring in September, one month later than that in Xinyang City. Five strains of genotype ⅠJE virus were isolated from mosquitoes in the two cities. Conclusions: From 2004 to 2025, the incidence of JE in Henan Province declined while the incidence in the ≥15 age group increased. The age distribution and temporal patterns of JE cases across regions were consistent with local antibody levels in the healthy population and the seasonal fluctuations of mosquito populations. It is necessary to strengthen surveillance and prediction and implement differentiated, precise prevention and control measures to reduce incidence.

PMID:42618481 | DOI:10.3760/cma.j.cn112338-20260211-00119