BMC Medicine最新文献

Background: Necrotising soft tissue infections (NSTI) are life-threatening conditions caused by diverse bacteria. Treatment strategies have remained largely universal and unchanged, and only modest improvements in patient outcomes have been observed. Emerging insights into NSTI pathogenesis may enable more targeted approaches. Because microbial aetiology is central to guiding appropriate therapy, we aimed to develop and externally validate machine learning models capable of predicting microbial aetiology using only data available at an early stage. In parallel, we explored whether similar models could predict selected clinical endpoints related to surgical management, patient handling, and organ support.

Methods: We used data from the INFECT study, an international multicentre prospective cohort investigating NSTI characteristics and pathogenesis. A total of 409 adults with surgically confirmed NSTI were enrolled between February 2013 and June 2017 from five Scandinavian hospitals. More than 700 clinical variables were collected from hospital admission to intensive care unit entry. Machine learning models were developed to predict the presence of Streptococcus pyogenes (GAS, Group A streptococcus) and five clinical endpoints: risk of amputation, size of skin defect, maximum skin defect size, length of intensive care (ICU) stay, and need for renal replacement therapy. Unsupervised variable selection was implemented, and Shapley Additive explanations were used for model interpretability. External validation employed a retrospective multicentre cohort of 216 NSTI patients treated in 11 Dutch hospitals between January 2013 and December 2017.

Results: Eight presurgical variables (age, diabetes, affected area, prior surgical intervention, and blood creatinine and haemoglobin concentrations) were sufficient for predicting GAS aetiology with high discriminatory power. Performance was good in both the development cohort (ROC-AUC 0.828; 95% CI 0.763-0.883) and the external validation cohort (ROC-AUC 0.758; 95% CI 0.696-0.821). Prediction of clinical endpoints related to surgical management, ICU stay, and organ support was unsuccessful.

Conclusions: We developed and externally validated a model predicting GAS aetiology in NSTI using presurgical data alone. Early identification of GAS may improve clinical handling and support tailored decisions on treatment and infection control, including management of close contacts and reduction of hospital transmission risk.

Background: Whole food plant-based diets exert anti-inflammatory properties and have been associated with clinical improvements in patients with autoimmune disorders. The underlying mechanisms remain poorly understood and functional insights into nutrient-host physiology cross-talks are urgently warranted. The present study investigated the effects of an isocaloric 8-week vegan diet (VD) intervention on whole blood count parameters and lymphoid composition in comparison to a meat-rich diet (MD).

Methods: We conducted a two-arm, monocentric randomized-controlled trial with healthy adults who were randomly allocated to either a MD or a VD for 8 consecutive weeks. Foods of animal origin were not permitted on the VD, whereas participants in the MD group were asked to consume at least 150 g of meat per day.

Results: Fifty-seven participants completed the study. At week 8, significant between-group differences were found for the white blood cell count (median (interquartile range): 5.17 (1.62) *10³/µL in the VD group vs. 5.39 (1.92) *10³/µL in the MD group, p = 0.029) and the lymphocyte count (1.80 ± 0.53 *10³/µL in the VD group vs. 2.06 (0.74) *10³/µL in the MD group, p = 0.049). This difference was driven by an increase in lymphocytes in MD group participants over the course of the study. Median change scores in platelets differed between VD and MD participants (- 21 (- 31) *10³/µL in the VD group vs. - 1.21 ± 28.37 *10³/µL in the MD group, p = 0.035) and so did the neutrophil change scores (- 0.17 (- 0.31) *10³/µL vs. 0.13 (0.50) *10³/µL, p = 0.034). Mixed models for repeated measures with a time-diet interaction as a fixed effect suggested that changes in white blood cells were driven by the diet factor alone (contrast: - 0.50 (95% CI: - 0.99-(- 0.01)), p = 0.046). Immunophenotyping results suggested significant between-group differences in CD3⁺ and CD8⁺ T-cells, and CD19⁺ B-cells after 8 weeks. CD19⁺ B-cells decreased significantly in the vegan group (214.77 ± 96.64 at baseline vs. 171.56 (102.73) cells/µL at week 8).

Conclusions: The present study suggests that a VD, in comparison to a MD, reduces the number of various immune cells even in healthy individuals. A VD may thus exert anti-inflammatory properties.

Trial registration: Registered at Deutsches Register Klinischer Studien: DRKS00031541.

Background: Clinical trials for multiple sclerosis (MS) disease-modifying treatments selectively enroll patients with favorable risk-benefit profiles. However, these therapies are often prescribed more broadly in clinical practice. We aimed to identify which patients are unlikely to benefit and may face substantial harm, and codify this into a data-driven framework for guiding real-world MS treatment decisions.

Methods: Systematic searches of PubMed and ClinicalTrials.gov identified 61 randomized, blinded phase 2b/3 trials with ≥ 100 adults per arm (all pediatric trials were included due to rarity), ≥ 48 weeks of treatment, and Expanded Disability Status Scale-based confirmed disability progression as an outcome. These trials enrolled 46,611 participants and contributed 91,787 patient-years. We extracted 80 baseline variables per trial arm and derived 30 additional features to reduce bias and train multivariable regression models. Model performance was validated using an independent, longitudinal real-world MS cohort. Infection-related mortality risk was estimated from national life tables and adjusted by treatment-specific hazard ratios.

Results: Baseline characteristics predicted both untreated progression and treatment efficacy. Therapeutic benefit increased with higher relapse rates and presence of enhancing lesions and declined with age and disease duration. Relapse rates in placebo arms declined across trial periods, mirrored by waning treatment efficacy on disability progression, which was confirmed in real-world data. In contrast, treatment-related morbidity and mortality increased with age, disability, and comorbidities. These opposing trends were integrated into a web-based personalized risk-benefit estimator.

Conclusions: Interpretable models offer a unified view of MS evolution and treatment effects. They show that the therapeutic risk-benefit ratio is dynamic, shaped by individual characteristics and predictable over time. The models project that initiating high-efficacy treatments early, followed by strategic de-escalation yields the best long-term outcomes. Critically, they extrapolate, and real-world data confirm that prescribing disease-modifying treatments to patients who would have been excluded from pivotal trials is more likely to cause harm than benefit. By enabling individualized, evidence-based decisions, this estimator can help clinicians deliver safer, more effective MS care worldwide.

Background: Retinal degenerative diseases represent a complex global health problem due to their significant impact on patients' daily lives and their highly heterogeneous pathogenesis, which challenges therapeutic development. Despite this complexity, many diseases, such as retinitis pigmentosa (RP) and age-related macular degeneration (AMD), share common features, including disrupted proteostasis, oxidative stress, and inflammatory responses, eventually leading to photoreceptor (PR) degeneration and vision loss. The inhibition of valosin-containing protein (VCP) has emerged as a promising mutation-independent therapeutic strategy for RP. However, clinical translation requires rigorous validation in models that closely reflect human retinal physiology.

Methods: Organotypic retinal explants from porcine, macaque, and human donors were placed in an in vitro culture setup and treated with ML240, a selective VCP inhibitor, delivered either as a free compound or encapsulated in mPEG_5kDa-cholane. Photoreceptor survival was assessed via TUNEL assay, outer nuclear layer (ONL) row quantification, and immunostaining. Retinal inflammation was evaluated by microglial staining. A dose-response study was performed to determine safety margins across species, and additional retinal markers were used to assess the preservation of non-photoreceptor retinal cell populations.

Results: Porcine retinal explants exhibited progressive photoreceptor degeneration under ex vivo conditions. Treatment with ML240, particularly when formulated with mPEG_5kDa-cholane, significantly reduced photoreceptor cell death and microglial activation. Macaque and human explants exhibited minimal to no signs of degeneration. Treatment did not affect morphological or histological features of the explant, demonstrating the safety of ML240 in the primate retina.

Conclusions: VCP inhibition via ML240 demonstrates an uncompromised safety profile in porcine, macaque, and human retinal explants. In addition, the neuroprotective activity of ML240 was evident in porcine tissue. Formulation with mPEG_5kDa-cholane enhances the overall performance of the compound, supporting its use for future clinical application as a mutation-independent therapeutic approach for retinal degenerative diseases.

Background: Fragmentation of healthcare delivery can disrupt the maternal care continuum and undermine effective coverage. In Mexico's segmented health system, institutional discontinuities may exacerbate inequities in access and quality. We examined the prevalence, determinants, and consequences of fragmented healthcare (FHC) for effective maternal healthcare coverage (EMHC) between 2009 and 2023.

Methods: We conducted a retrospective, repeated cross-sectional analysis using nationally representative data from the 2014, 2018, and 2023 ENADID surveys, including 71,874 women aged 12-54 with a recent live birth. EMHC was defined as a composite indicator encompassing adequate antenatal care (ANC), skilled or institutional delivery, timely postpartum care, and a complication-free puerperium. FHC was defined as receiving ANC and delivery care from different healthcare providers. Pooled multivariable regressions with survey fixed effects assessed the association between FHC and EMHC, adjusting for sociodemographic and contextual characteristics.

Results: Between 2009 and 2023, roughly one in six women experienced FHC, while only one in three achieved EMHC. Fragmentation was more frequent among women covered by publicly subsidized insurance (Seguro Popular or INSABI), Indigenous women, those living in rural areas, and women with higher obstetric risk. Receiving ANC from private providers tripled the odds of FHC compared with women covered by employment-based social security. Women exposed to FHC had a 4.7 percentage point lower probability of achieving EMHC-equivalent to a 20% reduction in the odds of effective coverage (aOR = 0.80; 95% CI: 0.69-0.91). This adverse effect was consistent across survey waves and most pronounced among Ministry of Health users.

Conclusions: Fragmented maternal healthcare trajectories substantially reduce the likelihood of effective coverage, disproportionately affecting socioeconomically and ethnically disadvantaged populations. The observed reduction in EMHC underscores that fragmentation is not merely a clinical or operational issue, but a structural challenge that requires reforms to improve the coordination of care. Strengthening integration across maternal care networks, ensuring interoperability of health information systems, and adopting continuity-based financing models are critical to improving coordination. Addressing FHC could prevent incomplete or unsafe care and accelerate progress toward universal health coverage. These findings offer actionable lessons for Mexico and other middle-income countries confronting health system fragmentation.

Background: Cardiovascular disease (CVD) remains the leading cause of death worldwide. While traditional risk factors are well-established, emerging evidence suggests shift work causing circadian rhythm disruption significantly contributes to CVD risk. This systematic review investigated molecular mechanisms linking circadian disruption with cardiovascular pathophysiology through in vivo models.

Methods: We systematically searched Medline, Embase, and Web of Science through February 2025. Studies employing genetic (clock gene knockouts/mutations) or environmental (light phase shift, sleep deprivation) models of circadian disruption in vivo were included. Meta-analyses were performed for key cardiovascular indicators, and certainty of evidence was evaluated using a modified GRADE approach.

Results: Among 9012 references, 34 studies met inclusion criteria. Following quality assessment for study design and reporting, 32 studies with low or moderate risk of bias were included in the synthesis. Meta-analyses revealed cardiac hypertrophy as the most robust finding, with high-certainty evidence for increased left ventricular mass-to-body weight ratio (LV/BW; SMD: 0.89, 95% CI: 0.38 to 1.39) and moderate-certainty evidence for increased cardiomyocyte size. These convergent organ and cellular-level findings, supported by elevated natriuretic peptides and pro-fibrotic markers, indicate circadian disruption contributes to pathological cardiac remodeling. Sensitivity analyses revealed low-certainty evidence for impaired systolic function, with significant reductions in ejection fraction (SMD: - 1.70, 95% CI: - 3.22 to - 0.17) and fractional shortening (SMD: - 1.60, 95% CI: - 2.71 to - 0.49). Low-certainty evidence was found for impaired endothelium-dependent vasorelaxation (SMD: - 2.72, 95% CI: - 4.90 to - 0.53) based on three genetic model studies with high heterogeneity and elevated triglyceride levels (SMD: 1.64, 95% CI: 0.07 to 3.21). Other markers showed very low-certainty evidence.

Conclusions: This systematic review improves mechanistic understanding of CVD development following circadian misalignment by demonstrating cardiac hypertrophy as a major pathophysiological consequence in animal models. Cardiac structural changes at organ and cellular levels, supported by biomarkers of pathological remodeling, indicate circadian disruption contributes to adverse cardiac remodeling. Future animal research should prioritize standardized protocols, sex-balanced designs, and environmental models replicating human shiftwork patterns. Substantial epidemiological gaps remain, warranting further investigation in shift workers.

Background: The Chinese new insurance policy for proprotein convertase subtilisin/kexin type 9 inhibitors (PCSK9i), which made it more affordable to patients from January 1, 2022, may change lipid-lowering therapy (LLT) pattern for acute coronary syndromes (ACS), but real-world data is still lacking. This study assessed to evaluate the impact of the reimbursement policy on LLT trends and its implications.

Methods: This cohort study is based on a large, ongoing, and prospective Chinese Cardiovascular Association Database-Chest Pain Center. Between January 1, 2021, and June 30, 2022, 789,829 patients with ACS from 4441 centers aged 18-80 years were included. Patients discharged after and before the policy were observational and control group, respectively. Propensity score matching (PSM) was performed at discharge, 1 month, and 3 months. Temporal trends in LLT are grouped into the 7 most common strategies: (1) high-intensity statin (HIS) monotherapy, (2) moderate-intensity statin (MIS) monotherapy, (3) MIS plus ezetimibe, (4) MIS plus PCSK9i, (5) MIS plus ezetimibe and PCSK9i, (6) ezetimibe monotherapy, and (7) PCSK9i monotherapy.

Results: A total of 789,829 (age 64 ± 10.7 years; females 37.84%) patients were included (268,089 in observational group). Temporally, MIS plus PCSK9i, MIS plus ezetimibe and PCSK9i, and PCSK9i monotherapy increased, ezetimibe monotherapy decreased, and statins keep relatively constant after the new policy. After PSM in observational group, MIS monotherapy slightly decreased (81.61% vs 89.29%, P < 0.0001) at discharge but increased in 1 month (84.54% vs 80.11%, P < 0.0001) and 3 months (82.97% vs 80.09%, P = 0.0013). MIS plus PCSK9i kept growing (0.86% vs 0.33%, P < 0.0001; 1.99% vs 0.80%, P < 0.0001; and 2.57% vs 0.60%, P < 0.0001, respectively). MIS plus ezetimibe and PCSK9i (0.15% vs 0.06%, P < 0.0001), and PCSK9i monotherapy (0.09% vs 0.03%, P < 0.0001) just increased at discharge. Low-density lipoprotein cholesterol (LDL-C) goal attainment rate increased at 3 months (15.73% vs 14.02%, P = 0.0283, before PSM; 15.73% vs 13.97%, P = 0.0306, after PSM). MIS plus PCSK9i and MIS monotherapy was associated with a higher 3-month LDL-C goal attainment rate.

Conclusions: The PCSK9i reimburse policy adjustment was associated with increased intensive LLT for ACS in China, especially more PCSK9i prescription, which modestly correlated to higher LDL-C goal attainment rate.

Background: Different patterns of non-communicable diseases (NCDs) in older adults may lead to distinct transitions of frailty progression and mortality risk. This study aimed to determine whether specific multimorbidity patterns of configurations (MCs) are associated with differences in frailty transitions and mortality risk.

Methods: Longitudinal data from 14,511 adults aged ≥ 60 years in the Lifelines cohort were analyzed, with median follow-up of 3.8 years for frailty and 5.6 years for mortality. Multimorbidity was assessed both by accumulation (≥ 2 NCDs) and by MCs previously identified through latent class analysis. Frailty was measured using a 32-item index and categorized as robust, pre-frail, or frail. Multistate Markov models estimated transitions between frailty states and to death; mixed-effects models assessed frailty average-changes over time.

Results: The prevalence of multimorbidity by NCD accumulation increased monotonically with frailty, whereas the distribution of MCs varied across frailty categories. The type of multimorbidity influenced the risk of transitioning between frailty states and to death. The 'Complex Treatment' group had the highest baseline frailty. The 'Major CVD & Vascular' and 'Heart & Vascular' configurations showed the steepest increases in frailty, while the 'Vascular' and 'Metabolic' groups presented lower frailty levels and a higher likelihood of recovery. Participants in the 'CVD & Vascular' group were more likely to transition from pre-frailty or frailty to death compared to other configurations.

Conclusions: Frailty transitions and mortality risk varied across MCs, identifying moderate- and high-risk profiles. MCs provide a more detailed interpretation of into how multimorbidity shapes frailty and mortality than simple disease accumulation.

Background: This study aimed to assess the prevalence of glucagon-like peptide-1 (GLP-1) receptor agonist use and interest in using medications for weight loss among adults (≥ 18 years) in Great Britain.

Methods: Nationally representative household survey, January-March 2025 (n = 5893). Participants were asked whether they had used medication in the past year to manage type 2 diabetes (excluding insulin), reduce the risk of heart disease, or support weight loss and, if so, whether they had used any of five specific GLP-1 or dual GLP-1/glucose-dependent insulinotropic polypeptide (GIP) receptor agonists. Those who had not used medication to support weight loss in the past year were asked how likely they would be to consider doing so in the next year. Estimates were reported stratified by participant characteristics and extrapolated to the national population.

Results: Overall, 2.9% [2.4-3.4%]- approximately 1.6 million adults-reported using a GLP-1 or GLP-1/GIP medication to support weight loss in the past year, with 1.7% [1.4-2.1%] (~ 910,000 adults) using them exclusively for this purpose. Of those who used them exclusively for weight loss, the majority (91.4% [85.6-97.2%]) reported using medications licensed for this purpose in Great Britain, most commonly Mounjaro (tirzepatide; 80.2% [71.9-88.6%]). Of those who had not used weight-loss medication in the past year, 6.5% [5.7-7.3%] (~ 3.3 million adults) expressed an interest in doing so in the next year. Use and interest were more prevalent among women, people in mid-life, and those reporting past-month psychological distress. Interest was also higher among people facing greater socioeconomic disadvantage.

Conclusions: In the first quarter of 2025, an estimated 4.9 million adults in Great Britain-nearly one in ten-either had recently used a GLP-1 or GLP-1/GIP medication to support weight loss or were interested in doing so in the near future. A substantial minority reported using a type of GLP-1 medication that was not licensed for weight management, suggesting off-label use. Interest was particularly high among less advantaged socioeconomic groups, while use was similar across groups, highlighting the importance of addressing equity in access. These findings underscore the need to monitor who is accessing these medications and to ensure their safe, appropriate, and equitable provision.

Background: Sodium-glucose co-transporter 2 inhibitors (SGLT2i), when combined with metformin (COMBI), offer multi-organ protective effects in patients with type 2 diabetes (T2D), particularly those at high risk of cardiovascular or renal complications. However, the underlying molecular mechanisms remain poorly understood.

Methods: We profiled 303 targeted serum metabolites in 1494 participants of the KORA study, including T2D patients treated with COMBI therapy, metformin monotherapy, or no glucose-lowering medication. Additionally, metabolomic profiling was quantified on seven tissues (plasma, liver, adrenal glands, adipose tissue, testis, lung, and cerebellum), and related hepatic transcripts were evaluated in 40 mice. Multivariable linear regression analyses, adjusted for age, sex, BMI, lifestyle, glycemic, and cardiovascular risk factors, were applied to human data; tissue-specific regression analyses were conducted for murine samples. Identified metabolites were further investigated using biochemical pathway analyses and literature review.

Results: COMBI therapy was associated with significant changes in metabolite profiles. In humans, 10 metabolites were significantly altered compared to metformin monotherapy. In mice, 82 altered metabolites were identified in plasma, 52 in liver, 30 in adrenal glands, 12 in adipose tissue, seven in testis, seven in lung, and six in cerebellum. COMBI therapy lowered threonine concentrations in both human serum and murine plasma but raised threonine, glycine, and urea cycle metabolites (citrulline, asymmetric dimethyl arginine (ADMA), and ornithine) in murine liver. This was accompanied by enhanced hepatic expression of Slc38a2, a threonine transporter gene. In humans, urea cycle metabolites correlated strongly with the fibrosis-4 index, a marker of liver fibrosis. Additionally, COMBI therapy elevated ketone body markers, such as hydroxybutyrylcarnitine, across murine liver, plasma, adrenal glands, adipose tissue, and testis.

Conclusions: COMBI therapy modulates amino acid metabolism, the urea cycle, and ketone body production, suggesting potential mechanisms underlying its protective effects against liver fibrosis and male subfertility. These findings provide novel insights into the systemic metabolic actions of COMBI therapy and highlight its translational potential to improve clinical outcomes in T2D patients.