Nevin Manimala

Sci Rep. 2025 Aug 23;15(1):30983. doi: 10.1038/s41598-025-16817-5.

ABSTRACT

Obesity and aging are key research topics in contemporary biomedical science. While studies have explored the effects of obesity on various health indicators, the precise mechanisms through which obesity may affect leukocyte telomere length (LTL)-and whether this impact contributes to accelerated immune cell senescence-remain unclear and warrant further investigation. In this study, we employed single nucleotide polymorphisms (SNPs) associated with four obesity indices-body mass index (BMI), body fat percentage (BFP), waist circumference (WC), and waist-hip ratio (WHR)-as instrumental variables (IVs) to assess the causal relationship between these indices and LTL through Mendelian randomization (MR) analysis. Additionally, we analyzed transcriptome sequencing data from peripheral blood mononuclear cells (PBMCs) across three groups: lean individuals, individuals with obesity before undergoing bariatric surgery, and individuals with obesity after surgery, and focus on the expression changes of cellular senescence and telomere dynamics related genes in PBMCs of individuals with obesity before and after weight loss intervention. The results showed a negative causal relationship between BMI (B=-0.04, P < 0.0001), BFP (B=-0.06, P < 0.0001) and LTL without being impacted by lipid profiles and T2D. The negative causal relationship between WC (B=-0.04, P < 0.0001) and LTL may be dependent on lipid levels, but not on T2D. WHR had no significant causal relationship (P > 0.05). Transcriptomic analysis further revealed that individuals with obesity had higher expression of cellular senescence-related genes such as ID2, LMNA, and TENT4B in PBMCs compared to lean individuals, with expression levels of these genes significantly decreasing after bariatric surgery. These findings underscore the detrimental impact of obesity on telomere attrition and immune cell senescence, highlighting the potential benefits of obesity management for slowing the biological process of cellular and immune aging.

PMID:40847086 | DOI:10.1038/s41598-025-16817-5

J Expo Sci Environ Epidemiol. 2025 Aug 22. doi: 10.1038/s41370-025-00798-8. Online ahead of print.

ABSTRACT

BACKGROUND: Perfluorooctane sulfonic acid (PFOS) is a legacy chemical, that while banned in some countries, is still found in various environmental media and in nearly all humans given its long half-life.

OBJECTIVE: We examined mean birth weight (BW) differences in relation to PFOS exposure biomarkers using systematic review methods.

METHODS: We fit a random effects model to obtain the overall pooled effect and for stratified analyses examining biomarker sample type and timing, study confidence, scaling factors, and country of study origin. We also conducted a meta-regression to assess the impact of gestational age and other factors on the overall pooled effect.

RESULTS: We found a 30-gram BW deficit (β = -30.3 g; 95%CI: -41.6, -18.9) with each ln-unit PFOS increase based on 53 studies identified in the systematic literature review. We detected BW deficits across all study confidence levels (β range: -27 to -37 g per ln-unit increase) with the largest deficit in the medium confidence grouping (β = -36.6 g; 95%CI: -56.3, -16.8). We did not see evidence of a gradient of BW deficits across biomarker sample timing (β range: -24 to -39 g per ln-unit increase), but the smallest deficit in our primary analyses was detected for the 18 early sample timing studies (β = -23.6 g; 95%CI: -38.7, -8.6). Robust deficits were also seen across various subgroups including by geographical region of study origin (e.g., Asian studies), more restrictive early biomarker sample collection, and post-partum samples (β range: -16.9 to -30.6 g). For meta-regression analyses, none of the investigated factors explained significant heterogeneity across studies.

IMPACT: We detected a statistically significant BW deficit of 30 grams per each ln-unit PFOS increase across all 53 studies in our meta-analysis; results were comparable in magnitude across study confidence, sample timing, and other strata. Unlike previous meta-analyses based on fewer studies, our results suggest that pregnancy hemodynamics do not fully explain the overall association. Characterization of the potential risk of developmental effects related to PFOS and other legacy chemicals will have important risk assessment and risk management ramifications in the future.

PMID:40847085 | DOI:10.1038/s41370-025-00798-8

Eur Radiol. 2025 Aug 22. doi: 10.1007/s00330-025-11924-3. Online ahead of print.

ABSTRACT

OBJECTIVE: Aimed to evaluate the potential of large language models (LLMs) in differentiating intra-axial primary brain tumors using structured magnetic resonance imaging (MRI) reports and compare their performance with radiologists.

MATERIALS AND METHODS: Structured reports of preoperative MRI findings from 137 surgically confirmed intra-axial primary brain tumors, including Glioblastoma (n = 77), Central Nervous System (CNS) Lymphoma (n = 22), Astrocytoma (n = 9), Oligodendroglioma (n = 9), and others (n = 20), were analyzed by multiple LLMs, including GPT-4, Claude-3-Opus, Claude-3-Sonnet, GPT-3.5, Llama-2-70B, Qwen1.5-72B, and Gemini-Pro-1.0. The models provided the top 5 differential diagnoses based on the preoperative MRI findings, and their top 1, 3, and 5 accuracies were compared with board-certified neuroradiologists’ interpretations of the actual preoperative MRI images.

RESULTS: Radiologists achieved top 1, 3, and 5 accuracies of 85.4%, 94.9%, and 94.9%, respectively. Among the LLMs, GPT-4 performed best with top 1, 3, and 5 accuracies of 65.7%, 84.7%, and 90.5%, respectively. Notably, GPT-4’s top 3 accuracy of 84.7% approached the radiologists’ top 1 accuracy of 85.4%. Other LLMs showed varying performance levels, with average accuracies ranging from 62.3% to 75.9%. LLMs demonstrated high accuracy for Glioblastoma but struggled with CNS Lymphoma and other less common tumors, particularly in top 1 accuracy.

CONCLUSION: LLMs show promise as assistive tools for differentiating intra-axial primary brain tumors using structured MRI reports. However, a significant gap remains between their performance and that of board-certified neuroradiologists interpreting actual images. The choice of LLM and tumor type significantly influences the results.

KEY POINTS: Question How do Large Language Models (LLM) perform when differentiating complex intra-axial primary brain tumors from structured MRI reports compared to radiologists interpreting images? Findings Radiologists outperformed all tested LLMs in diagnostic accuracy. The best model, GPT-4, showed promise but lagged considerably behind radiologists, particularly for less common tumors. Clinical relevance LLMs show potential as assistive tools for generating differential diagnoses from structured MRI reports, particularly for non-specialists, but they cannot currently replace the nuanced diagnostic expertise of a board-certified radiologist interpreting the primary image data.

PMID:40847080 | DOI:10.1007/s00330-025-11924-3

Eur Radiol. 2025 Aug 22. doi: 10.1007/s00330-025-11938-x. Online ahead of print.

ABSTRACT

OBJECTIVE: We evaluated the value of EC as a major feature for hepatocellular carcinoma (HCC) and its implications on simplifying LI-RADS v2018: by either excluding EC (LI-RADS[EC(-)]), or treating EC as equivalent to other major features (LI-RADS[EC(+)]).

MATERIALS AND METHODS: This retrospective study included patients who underwent preoperative MRI within 1 month of surgery. Two radiologists independently assigned LI-RADS categories, with consensus resolution. The association between EC and HCC was evaluated by diagnostic odds ratio (DOR). Sensitivity and specificity of LI-RADS[EC(-)] and LI-RADS[EC(+)] were compared to LI-RADS v2018 using generalized estimating equations.

RESULTS: Among 735 patients (median [interquartile range] age, 60 [54-66] years, 562 men) with 986 observations, 150 (20.4%) underwent extracellular contrast (ECA)-MRI and 585 (79.6%) underwent hepatobiliary contrast (HBA)-MRI. EC was significantly associated with HCC on both ECA-MRI (DOR = 16.6, p < 0.001) and HBA-MRI (DOR = 9.3, p < 0.001). LI-RADS[EC(+)] strategy demonstrated significantly higher sensitivity than LI-RADS v2018 on both ECA-MRI (81.3% vs. 78.4%, p = 0.047) and HBA-MRI (74.2% vs. 72.7%; p = 0.005), without significant differences in specificity (96.7% vs. 96.7%; p > 0.99 for ECA-MRI and 91.7% vs. 92.1%; p = 0.26 for HBA-MRI). However, LI-RADS[EC(-)] yielded a lower sensitivity (74.8% vs. 78.4%; p = 0.03 for ECA-MRI and 72.3% vs. 72.7%; p = 0.17 for HBA-MRI) without a gain in specificity.

CONCLUSION: EC is strongly associated with HCC and critical for maintaining the sensitivity of the LI-RADS v2018. Treating EC as equivalent to other major features may simplify LI-RADS and improve sensitivity, particularly on ECA-MRI, where the increase in sensitivity was more pronounced than on HBA-MRI.

KEY POINTS: Question Can LIRADS v2018 be simplified by considering enhancing capsule (EC) as equivalent to other major features (i.e., non-peripheral washout), regardless of observation size? Findings Treating EC as equivalent to other major features yielded higher sensitivity than LIRADS v2018 on both extracellular- and hepatobiliary contrast-enhanced MRI, without significantly compromising specificity. Clinical relevance EC is a key feature for hepatocellular carcinoma diagnosis. The strategy of treating EC as equivalent to other major features can be used to increase diagnostic sensitivity as well as simplify LI-RADS v2018, particularly on extracellular contrast-enhanced MRI.

PMID:40847079 | DOI:10.1007/s00330-025-11938-x

Mamm Genome. 2025 Aug 22. doi: 10.1007/s00335-025-10154-8. Online ahead of print.

ABSTRACT

The relationship of inflammatory cytokines with the subtypes and prognosis of stroke is not fully understood. Mendelian randomization (MR) was used to evaluate the bidirectional relationship of inflammatory cytokines with stroke subtype (both ischemic and hemorrhagic), and functional outcome of ischemic stroke (modified Rankin Scale score), using databases from Genome-wide association studies, the GISCOME study, the UK Biobank, deCODE, and ONTIME. Colocalization analysis was conducted to determine whether cytokines and stroke subtypes had associations with the same single-nucleotide polymorphism (SNP). Meta-analysis of MR was performed to prove the robustness of the causal relationship between cytokines and stroke subtypes. In addition, both two-step MR analysis and multivariate MR were utilized in mediation analysis to ascertain whether inflammatory cytokines affected stroke subtypes through their regulation of risk factors of cerebrovascular diseases. MR revealed that the genetic prediction of circulating fibroblast growth factor 5 (FGF5) was associated with an increased risk of ischemic stroke and intracranial hemorrhage, but not with the functional outcome of ischemic stroke. Colocalization analysis demonstrated that the association of FGF5 with ischemic stroke and intracranial hemorrhage was driven by the same SNPs. Meta-analyses supported the causal relationship of FGF5 with ischemic stroke and intracranial hemorrhage. Mediation analyses revealed that both essential hypertension and atrial fibrillation mediate the increased risk of ischemic stroke and intracranial hemorrhage due to FGF5. Inflammatory cytokines are associated with stroke and risk factors of cerebrovascular diseases. A high level of circulating fibroblast growth factor 5 is a potential risk factor for stroke.

PMID:40847077 | DOI:10.1007/s00335-025-10154-8

Int J Obes (Lond). 2025 Aug 22. doi: 10.1038/s41366-025-01862-x. Online ahead of print.

ABSTRACT

BACKGROUND: The incidence of endometrial cancer has been rising sharply among younger generations, paralleling the growing obesity epidemic in this age group. Abdominal obesity is currently being investigated as an indicator of adiposity and cancer risk, and its prevalence is increasing in young women. This study aimed to examine whether cumulative abdominal obesity exposure in young women was associated with the development of endometrial cancer.

METHODS: We used data from the South Korean National Health Insurance Service for women aged 20-39 years who had completed four consecutive annual health examinations between 2009 and 2012 and had no history of cancer at baseline. Participants were categorized into five groups based on the number of abdominal obesity exposures (waist circumference ≥ 85 cm). Exposure numbers ranged from 0 to 4, indicating the frequency of abdominal obesity across the four health examinations over 4 years. The primary outcome was newly diagnosed endometrial cancer, which was monitored until 2020, with a follow-up period of 7.12 years.

RESULTS: Among the 445,791 young women (mean [SD] age 30.82 [4.55] years), 302 (mean [SD], 32.79 [4.53] years) developed endometrial cancer. The cumulative incidence of endometrial cancer differed significantly according to the number of abdominal obesity exposures (log-rank test, P < .001). The incidence of endometrial cancer has progressively increased with abdominal obesity exposure. The multivariable-adjusted HRs for incident endometrial cancer were 1.480 (95% CI, 0.970-2.258), 2.361 (95% CI, 1.391-4.008), 4.114 (95% CI, 2.546-6.647), and 6.215 (95% CI, 4.250-9.088) for participants with exposure numbers of 1-4, respectively, compared with those with an exposure number of 0.

CONCLUSION: In this population-based nationwide cohort study of young women, we observed a progressive increase in the risk of endometrial cancer with cumulative abdominal obesity exposure.

PMID:40847073 | DOI:10.1038/s41366-025-01862-x

Int J Obes (Lond). 2025 Aug 22. doi: 10.1038/s41366-025-01877-4. Online ahead of print.

ABSTRACT

OBJECTIVES: The association between obesity and cholelithiasis has been identified. However, the causal relationship between age-specific childhood obesity and adult cholelithiasis remains unclear. In addition, the biological basis for the association between childhood obesity and adult cholelithiasis is poorly understood, which poses a challenge for preventing adult cholelithiasis in specific biological pathways.

METHODS: Summary statistics of genome-wide association studies (GWASs) of childhood age-specific body mass index (BMI) at 12 time points and adult cholelithiasis derived from FinnGen were used in this study, with the former covering data from birth to 8 years. Linkage disequilibrium score regression (LDSC) analyses were used to assess the genetic correlations of age-specific childhood BMI to cholelithiasis. Two-sample Mendelian randomization (MR) and multivariable Mendelian randomization (MVMR) analyses were utilized to explore the causal associations. As downstream analyses, summary-based Mendelian randomization (SMR) analyses, transcriptome-wide association studies (TWAS), and Bayesian colocalization were conducted to discover the shared transcriptomic signals. The GWAS summary statistics of cholelithiasis from the UK Biobank were used for sensitivity analyses.

RESULTS: LDSC analyses revealed significant genetic correlations between 11 age-specific childhood BMIs and adult cholelithiasis (except for birth BMI). Two-sample MR and MVMR analyses indicated causal relationships between birth BMI and BMI at 8 months, 1.5 years, 7 years, and 8 years after birth and adult cholelithiasis. SMR, TWAS, and colocalization analyses identified MLXIPL as the strongest overlapping signal between age-specific BMI and adult cholelithiasis.

CONCLUSION: This study provides new evidence on the relationships between childhood obesity and adult cholelithiasis, highlighting the role of early intervention for obesity in childhood at key time points. MLXIPL gene expression was identified as a potential biological pathway, suggesting potential therapeutic targets and precise intervention strategies for childhood obesity and adult cholelithiasis.

PMID:40847070 | DOI:10.1038/s41366-025-01877-4

Int J Obes (Lond). 2025 Aug 22. doi: 10.1038/s41366-025-01889-0. Online ahead of print.

ABSTRACT

The primary aim was to develop and assess the performance and applicability of different models utilizing sensor data to determine dietary adherence, specifically within the context of intermittent fasting. Our approach utilized time-series data from two completed human trials, which included continuous glucose monitoring, acceleration data, and food diaries, and a synthetic data set. Machine learning models achieved an average F1-score of 0.88 in distinguishing between fasting and non-fasting times, indicating a high level of reliability in classifying fasting states. The Hutchison Heuristic statistical method, while more moderate in performance, proved to be robust across different cohorts, including individuals with and without type 1 diabetes. A dashboard was developed to visualize results efficiently and in a user-friendly manner. The findings highlight the effectiveness of using sensor data, combined with advanced statistical and machine learning approaches, to passively evaluate dietary adherence in an intermittent fasting context.

PMID:40847068 | DOI:10.1038/s41366-025-01889-0

Sci Rep. 2025 Aug 22;15(1):30964. doi: 10.1038/s41598-025-16531-2.

ABSTRACT

Statistical Process Control is essential for ensuring process stability and detecting variations in a production environment. This study introduces a control chart based on the Exponentially Weighted Moving Average (EWMA) that uses an adaptive sample size. The proposed approach enhances shift detection by dynamically adjusting the sample size in response to changes in process variation. Extensive Monte Carlo simulations were performed to assess the performance of the proposed control chart, focusing on metrics such as the Average Run Length (ARL) and the Standard Deviation of Run Length (SDRL). The findings show that the new chart surpasses both the Fixed Sample Size EWMA (FEWMA) and the Variable Sample Size EWMA charts, particularly in detecting small to moderate shifts in the process. This approach strikes a balance between detection sensitivity and computational efficiency, enabling prompt identification of process changes while maintaining robustness during in-control conditions. To illustrate its practical applicability, a real-world dataset was analyzed, demonstrating the effectiveness of the proposed method in actual process monitoring scenarios.

PMID:40847067 | DOI:10.1038/s41598-025-16531-2

Sci Rep. 2025 Aug 22;15(1):30927. doi: 10.1038/s41598-025-16526-z.

ABSTRACT

High-dimensional longitudinal data present significant analytical challenges due to intricate within-subject correlations and an overwhelming ratio of predictors to observations. To address these challenges, we introduce Mixed-Effect Gradient Boosting (MEGB), a novel R package that synergises gradient boosting with mixed-effects modelling to simultaneously account for population-level fixed effects and subject-specific random variability. MEGB provides a unified framework for analysing repeated measures data that accommodates complex covariance structures while harnessing gradient boosting’s inherent regularisation for robust feature selection and prediction. In comprehensive simulations spanning linear and nonlinear data-generating processes, MEGB achieved 35-76% lower mean squared error (MSE) compared to state-of-the-art alternatives like Mixed-Effect Random Forests (MERF) and REEMForest, while maintaining 55-70% true positive rates for variable selection in ultra-high-dimensional regimes $(p=2000)$ . Demonstrating practical utility, we applied MEGB to maternal cell-free plasma RNA data $(n=12$ subjects, $p=33,297$ transcripts), where it identified 9 key placental transcripts driving fetal RNA dynamics across pregnancy trimesters.

PMID:40847064 | DOI:10.1038/s41598-025-16526-z