Tracheostomy is frequently performed in critically ill patients requiring prolonged mechanical ventilation, but evidence regarding its prognostic impact in infectious diseases remains limited. This retrospective cohort study analyzed 607 patients who underwent tracheostomy at a specialized infectious disease hospital in Brazil between May 2020 and April 2025. Demographic data, comorbidities, severity scores, perioperative variables, and clinical outcomes were extracted from medical records, and multivariate logistic regression identified factors independently associated with in-hospital mortality. Overall hospital mortality was 64.6%. Age over 60 years, diabetes mellitus, higher SAPS 3 score, coagulopathy, need for hemodialysis, and a PaO₂/FiO₂ ratio below 200 on the day of tracheostomy were independently associated with death. In a sensitivity analysis excluding patients with COVID-19, HIV infection (analyzed as an underlying comorbidity) was associated with increased mortality, while the use of combined antiretroviral therapy was associated with improved survival; these results should be interpreted as exploratory. Overall, outcomes after tracheostomy in patients with infectious diseases appear to be driven primarily by systemic severity and organ dysfunction rather than by infectious etiology, underscoring the importance of individualized, multiparametric prognostic assessment.

Early identification of individuals at risk of type 2 diabetes is important in young and middle-aged adults. Although the homeostatic model assessment of insulin resistance (HOMA-IR) is widely used, its limited clinical practicality has increased interest in the triglyceride-glucose (TyG) index as a simpler surrogate marker of insulin resistance. However, the predictive values of longitudinal changes in these indices remain unclear. We analyzed data from 271,592 Korean adults aged 18-49 years who underwent at least two health screening examinations. Changes in the TyG index and HOMA-IR were categorized into sex-specific quintiles. Incident type 2 diabetes was defined as fasting glucose ≥ 126 mg/dL, HbA1c ≥ 6.5%, or initiation of antidiabetic medication. Hazard ratios were estimated using the Cox proportional hazard models. During the follow-up period, 3,841 men and 1,069 women developed type 2 diabetes. Greater increases in the TyG index were consistently associated with a higher risk of diabetes across quintiles in both sexes. The associations with HOMA-IR changes were more variable across analytical models, particularly in the lower quintiles. These findings suggest that monitoring changes in both indices may provide additional information for diabetes risk assessment, with the TyG index showing more consistent associations across models in our study. Further studies in diverse populations are needed to confirm these observations.

Efsubaglutide alfa, a long-acting humanised GLP-1 receptor agonist, was evaluated as add-on therapy to metformin in adults with type 2 diabetes. In an operationally seamless adaptive phase 2b/3, randomised, double-blind, placebo-controlled trial, phase 2b randomised participants 1:1:1 to Efsubaglutide 1 mg, 3 mg, or placebo for 12 weeks; based on phase 2b efficacy and safety, an independent committee selected 3 mg as the recommended phase 3 dose. Phase 3 randomised new participants 1:1 to Efsubaglutide 3 mg or placebo for 24-week double-blind treatment, followed by a 28-week open-label extension. At week 12 in phase 2b, HbA1c decreased by 1.10% with 1 mg and 1.43% with 3 mg (both p < 0.0001). At week 24 in phase 3, HbA1c decreased by 1.80% with Efsubaglutide 3 mg versus 0.74% with placebo (estimated treatment difference, -1.06%; p < 0.001). Adverse events were predominantly mild to moderate gastrointestinal events. Trial registration: ClinicalTrials.gov NCT04998032.

Meteorin-like protein (Metrnl) is a newly discovered adipokine that affects lipid and glucose metabolism and insulin signaling. The present study aimed to measure serum Metrnl levels in nonalcoholic fatty liver disease (NAFLD) patients with and without type 2 diabetes mellitus (T2DM) and investigate its association with carotid intima-media thickness (cIMT), a clinical marker of atherosclerosis.

Patients with NAFLD (n = 66), patients with NAFLD+T2DM (n = 66), and healthy individuals (n = 65) were included in this case-control study. Serum levels of Metrnl and adiponectin were determined via ELISA. cIMT and liver stiffness (LS) were also assessed in all participants.

Compared with controls, patients with NAFLD+T2DM had lower levels of adiponectin, which was accompanied by a lower level of Metrnl (P < 0.001). Furthermore, multinomial logistic regression revealed that serum Metrnl was associated with NAFLD and NAFLD+T2DM. This association remained statistically significant even after adjustment for either body mass index (BMI) or Homeostasis Model Assessment of Insulin Resistance (HOMA-IR). Pearson's correlation analysis in the patient group revealed significant inverse correlation between Metrnl levels and BMI, cIMT, insulin, and HOMA-IR. After adjustment for correlated factors in the patient group, multiple linear regression showed that multiple linear regression showed that BMI, carotid IMT, and HOMA-IR were independently associated with serum Metrnl levels. Moreover, Metrnl levels were lower in individuals with cIMT≥ 0.8 mm compared to those with cIMT< 0.8 mm.

The results of the present study indicated that the relationship between decreased Metrnl levels and NAFLD and T2DM could be related to obesity and IR. However, further studies are necessary to establish this concept.

Remnant cholesterol (RC) is a known contributor to cardiovascular disease, but its role in gestational diabetes mellitus (GDM) remains insufficiently characterized.

To examine the association between first-trimester RC levels and GDM risk, and to quantify the mediating effects of prepregnancy body mass index (BMI) and insulin sensitivity.

This prospective cohort study included 877 pregnant women in Southeastern China. RC was calculated from first-trimester fasting lipid profiles. GDM was diagnosed via a 75 g oral glucose tolerance test at 24 to 28 weeks. Binary logistic regression, restricted cubic splines, and serial mediation models were employed for analysis.

Among all participants, 160 (18.2%) developed GDM. After full adjustment for confounders, women in the highest quartile of first-trimester RC had a significantly higher risk of GDM (odds ratio = 2.850, 95% CI: 1.679-4.836) compared with those in the lowest quartile. A significant positive dose-response relationship was observed (P for overall < .001). Serial mediation analysis indicated that the association between RC and GDM was partially mediated through the sequential pathway of prepregnancy BMI → Matsuda index, accounting for 13.50% of the total effect. RC also showed modest predictive performance for GDM (area under the curve = 0.673) compared with conventional lipid parameters.

Elevated first-trimester RC is independently associated with increased GDM risk. This association is partially mediated by prepregnancy BMI and subsequent insulin resistance. Measuring RC in early pregnancy may improve the identification of women at high risk for GDM.

Diabetic cardiomyopathy is driven by oxidative stress, impaired myocardial enzyme function, and inflammatory/apoptotic signaling. Irisin has emerged as a cardiometabolic regulator, but its integrated effects on cardiac redox status, enzyme activities, and injury biomarkers in diabetes remain incompletely understood.

Diabetes-like cardiometabolic injury was induced in male Wistar rats by fructose pre-treatment followed by low-dose streptozotocin (40 mg/kg, i.p.). Diabetic rats were treated with irisin (100 or 500 μg/kg/day, i.p.) or metformin (200 mg/kg/day, i.p.) for 21 days. Cardiac oxidative stress markers (MDA and GSH), antioxidant enzyme activities (SOD, CAT, GPx, and GST), ATPase activities (Na⁺/K⁺-ATPase, Ca²⁺/Mg²⁺-ATPase, and Mg²⁺-ATPase), phosphatase activities (ALP and ACP), and cardiac p53 and VCAM-1 mRNA expression were assessed. Serum injury biomarkers (CK-MB, cTnI, cTnT, and NT-proBNP) and histopathological changes were also evaluated.

Cardiac MDA was higher, and GSH was lower in diabetic rats (both p < 0.001) with marked reductions in SOD, CAT, GPx, and GST activities (p < 0.001). Diabetes also inhibited the activities of Na + /K + -ATPase, Ca 2+ /Mg 2+ -ATPase, and Mg 2+ -ATPase (p <0.001) and reduced ALP and ACP (p <0.001). The mRNA levels of cardiac p53 and VCAM-1 were significantly elevated (p < 0.001), along with serum CK-MB, cTnI, cTnT, and NT-proBNP (p < 0.001). At 100 and 500 μg/kg, irisin significantly decreased MDA content and restored antioxidant enzymes, ATPases, phosphatases (all p < 0.001 vs diabetic control), and was associated with reduced p53 and VCAM-1 expression (p < 0.001); it also reduced CK-MB, troponins, and NT-proBNP levels (p < 0.001) more effectively at a higher tested dose (500 μg/kg). The histology results showed better myocardial architecture with irisin, similar to metformin.

Irisin produced marked improvement in biochemical and histological indices of diabetes-induced myocardial injury and was associated with improved redox balance, reduced p53 and VCAM-1 mRNA expression, and lower circulating cardiac injury biomarkers. These findings support a descriptive association between irisin treatment and attenuation of diabetes-related myocardial injury, while further protein-level and pathway-focused studies are needed to confirm the underlying mechanisms.

While classical diabetic foot risk factors are well established, their temporal progression remains insufficiently understood, particularly for deformities. Therefore, we aimed to analyze trends in selected diabetic foot risk factors and to design a practical monitoring model for long-term clinical use.

From 51,001 routine foot examinations at tertiary diabetes clinic (1998-2024), we included 14,436 screenings from 3,049 patients with complete data. Using generalized estimating equations, we modelled the temporal trends in the prevalence of six diabetes-related complications.

Loss of Protective Sensation (LOPS) significantly increased the odds (adjusted for age at diagnosis, sex, diabetes type, and duration) of five complications: fat pad atrophy (OR = 2.12), toenail deformity (OR = 2.44), toe deformities (OR = 2.57), callus (OR = 3.66), and xerosis (OR = 2.27). Toenail deformity was the most prevalent complication, while fat pad atrophy showed the steepest relative increase over time. Female sex was a risk factor for specific deformities and fat pad atrophy but protective against xerosis (OR = 0.76).

The developed models provide clinically actionable risk trajectories, revealing distinct patterns by complication type, LOPS status, and demographic factors. These findings can directly support targeted screening protocols and inform resource allocation.

In the EXSCEL trial, exenatide did not reduce major adverse cardiovascular events (MACE), but heterogeneity of benefit and the role of cardiac biomarkers remain uncertain. We evaluated the prognostic value of baseline and 1-year changes in N-terminal pro B-type natriuretic peptide (NT-proBNP) and high-sensitivity cardiac troponin I (cTnI), and whether baseline biomarker concentrations modified exenatide effects.

EXSCEL randomized 14,752 adults with type 2 diabetes to exenatide 2 mg weekly (EQW) or placebo. In a biomarker cohort, 4,292 participants had serial NT-proBNP or cTnI at baseline and 1 year. Biomarkers were log transformed and Cox models related baseline concentrations and 1-year change to MACE, all-cause mortality (ACM), cardiovascular (CV) death, hospitalization for heart failure (hHF), adjusting for clinical covariates and the alternate biomarker. Treatment interaction was tested with biomarker by treatment terms.

Over median 1,480 days follow-up, 529 MACE, 310 all cause deaths, 193 CV deaths, and 157 hHF events occurred. Baseline NT-proBNP was strongly prognostic (adjusted HR per 1 integer unit 1.63 for MACE, 1.85 for ACM, 2.17 for CV death, and 2.17 for hHF; all p<0.001). Baseline cTnI was also prognostic with a nonlinear pattern, with risk rising mainly above the median. Per SD rise in NT-proBNP over 1 year predicted later MACE (HR 1.85) and CV death (HR 2.81; both p<0.001). Baseline NT-proBNP didn't modify treatment effects. Baseline cTnI didn't modify EQW treatment effect on MACE but lower rates of CV deaths and hHF with EQW were observed at higher cTnI concentrations.

NT-proBNP and cTnI were strong prognostic markers of adverse outcomes in patients with type 2 diabetes and their 1-year increases signaled higher subsequent risk. Baseline cTnI may mark heterogeneity of EQW response, but mortality interactions are hypothesis generating and require confirmation.

Inflammatory cytokines, particularly interleukin-1β (IL-1β), have a main role in diabetic nephropathy (DN) pathogenesis. Genetic variations within the IL1B promoter region may modulate cytokine expression and affect vulnerability to renal injury in type 2 diabetes mellitus (T2DM). We aimed to assess the correlation between the IL1B (-511C/T;rs16944) polymorphism, IL-1β serum levels, and renal function among Egyptian cases with T2DM.

This study was conducted on 150 subjects, who were divided into 50 T2DM cases with DN, 50 T2DM patients without DN, and 50 healthy control subjects recruited from Mansoura University Hospital. Genotyping of IL1B (-511C/T) was conducted by TaqMan real-time PCR, and serum IL-1β concentrations were quantified by ELISA. Clinical and biochemical parameters, including HbA1c, serum creatinine (Ser Cr), urinary albumin, and estimated glomerular filtration rate (eGFR), were analyzed.

The TT genotype (GT) was significantly more prevalent in DN cases (76%) than in diabetic (48%) and control (44%) groups (p = 0.008). TT carriers exhibited higher IL-1β levels (18.3 ± 13.5 pg/mL), HbA1c (7.6 ± 2.5%), serum creatinine (1.94 ± 1.3 mg/dL), and urinary albumin excretion, alongside lower eGFR (70.3 ± 22.9 mL/min/1.73m²) compared with CT/CC genotypes (p < 0.05).

The IL1B (-511T) allele is accompanied by enhanced IL-1β secration, poorer glycaemic control, and renal impairment in T2DM, suggesting a genetic predisposition to inflammation-driven DN. Screening for this variant may aid early risk stratification and personalized therapeutic targeting of IL-1β pathways in diabetic kidney disease (DKD).

Accurate disease prediction using clinical datasets is essential for improving early diagnosis and clinical decision-support systems; however, many existing deep learning approaches are disease-specific, computationally intensive, and difficult to generalize across heterogeneous biomedical datasets. This study addresses this challenge by proposing a unified and dataset-aware deep learning framework that enables accurate and interpretable disease prediction across diverse clinical datasets. The framework adopts a modular architecture that selects appropriate models based on dataset characteristics such as feature dimensionality, sample size, and class imbalance. It integrates multiple deep learning architectures, including MLP, one-dimensional CNN, FT-Transformer, autoencoder-based classifiers, and ensemble strategies. Robust preprocessing, fold-safe feature selection, and nested cross-validation are incorporated to ensure reliable performance evaluation. The framework is evaluated on three heterogeneous benchmark datasets: the UCI Heart Disease dataset (303 samples, 13 clinical features), the PIMA Indians Diabetes dataset (768 samples, 8 metabolic features), and the Parkinson's disease voice dataset (195 recordings, 22 acoustic features). Experimental results demonstrate competitive predictive performance relative to classical baselines across the diverse tasks. The FT-Transformer + autoencoder ensemble achieved an AUC of 0.8980 (±0.0483) for heart disease prediction, while the CNN + Autoencoder ensemble obtained an AUC of 0.8451 (±0.0270) for diabetes classification. For Parkinson's disease detection, the MLP achieved an AUC of 0.7538 with perfect specificity. Overall, all models achieved AUC values comparable to ML baselines. The study contributes a scalable and interpretable deep learning framework that improves reliability, generalization, and practical applicability for multi-disease prediction in real-world healthcare environments.