Impact of first and further decompensation in patients with compensated ACLD due to MASLD.

Covariate measurement error is an important problem in survival analysis, which has been well studied under the Cox proportional hazards model. However, measurement error effects have been rarely addressed under the Aalen's additive hazards model, and there is a lack of methods to correct for error effects. In recent years, the Aalen's additive hazards model has been increasingly used in causal mediation analysis. Although the longitudinal mediator is frequently measured with uncertainty, the issue of measurement error in the mediator has received little attention. In this article, we study the general problem of covariate measurement error under theAalen's additive hazards model and propose a measurement error correction strategy. We then extend the proposed method to causal mediation analysis in the survival setting with an error-prone longitudinal mediator. Corrected estimation of the direct and indirect effects is obtained. The performance of the proposed method is assessed in numerical studies.

Conditional inference for ultrahigh-dimensional additive hazards model

A recent study has suggested that plastics may contain more than 16,000 chemicals, including additives, processing aids, starting substances, intermediates and Non-Intentionally Added Substances. Plastic chemicals are released throughout the plastic life cycle, from production, use, disposal and recycling. Most of these chemicals have not been studied for potential hazardous properties for humans and in the environment. To refine the risk assessment of these leachable chemicals, additional hazard data are needed. The PlasticLeach project within the EU co-funded Partnership for the Assessment of Risks from Chemicals (PARC) aims to address this data gap by screening several plastic products in daily use. Leachates will be prepared from a number of these plastic items, and these chemical mixtures will be further tested using several test guideline compliant assays and New Approach Methodologies covering both human health and environmental endpoints. The most toxic leachates will be characterized using a non-targeted analysis pipeline to identify chemicals in the leachate. When single chemicals of concern are identified, these will be further tested to determine hazardous properties and identify the respective potency factors to better understand their specific hazard profiles. A tiered approach for hazard testing will be followed. The experimental work will be complemented by in silico toxicological profiling, using publicly available toxicity databases and tools, including Artificial Intelligence tools that cover both human and environmental endpoints. A comprehensive array of endpoints, including cytotoxicity, endocrine disruption, genotoxicity, immunotoxicity, reproductive toxicity and effects related to ecotoxicity will be evaluated. In this paper, we outline the plastic products to be tested and the battery of assays that will be used to identify hazards relevant to both human health and the environment. Data generated from in silico , in vitro , and in vivo approaches will be reported using standardized formats, stored within a centralized repository, and harmonized to adhere to the FAIR data principles (Findable, Accessible, Interoperable, and Reusable). This integrated strategy will not only advance our understanding of the risks associated with plastic-derived chemicals but will also provide critical support for regulatory decision-making and facilitate the development of safer, and more ecofriendly plastic materials in the future.

The Central Sulawesi government has a Sustainable Development Goals (SDGs) target for 2020-2024, which sets the maternal mortality rate below 70/100,000 KH. However, in 2018-2022, the maternal mortality rate fluctuated by 128/100,000 KH. One of the factors causing maternal mortality is the duration of the labor process. The factors that are thought to have an influence on the duration of labor are gestational age, maternal age, baby height, parity, and hemoglobin levels. Therefore, this study aims to see what modeling and factors affect the duration of birth using Lin-Ying additive hazard regression analysis. Data were obtained from the medical records of normal deliveries between January and December 2023 at Anutapura Palu Hospital. The results showed that the factors that affect the duration of birth are preterm gestational age, aterm gestational age, maternal age 20-35 years, primigravida mothers, multigravida mothers, and mothers who are not anemic. A limitation of this study is the relatively short data collection period of one year, which may not capture variations or trends in labor outcomes over time.

BackgroundSleep duration has a well-established effect on mental health and well-being, with durations of 7 to 9 hours being the general recommendation. Here, we analyze the significance of sleep patterns and find that a consistent routine reduces the risk of developing mental disorders far more than simply ensuring a certain average sleep duration.MethodsWe analyzed the sleep behavior of 100,000 adults for one week using motion data from wrist-worn devices. We modeled sleep behavior using multivariate generalized additive Cox proportional hazard models, incorporating a smooth 2D interaction effect of sleep duration and routine sleep hours. We calculated C-statistics and E-values to evaluate model performance and assess the robustness against hidden confounders. We also stratified analyses by age and gender.ResultsMost participants slept for 7 to 9 hours as recommended, yet they consistently only slept during the same 4.8 hours each night. We found that an average sleep duration around 8 hours minimizes the risk of future mental disorders—but only if integrated into a rigorous sleep routine spanning at least the same 7 hours each night. Our study provides evidence that adopting such sleep behavior could reduce the population incidence rate of mental disorders by 23% (HR: 0.79, p<0.0001, for the average participant). The models showed a strong fit (C-statistics: 0.63), robustness to hidden confounders (E-value: 1.8), and stability under age- and gender-based stratification. We identified weekend behavior as a frequent reason for low sleep routines, with over 25% of the population disrupting their weekly sleep routine during weekend nights—raising the risk of future mental disorders by 10%.ConclusionsOur results suggest that maintaining a consistent sleep routine is more important for mental health than sleep duration alone. Socially disadvantaged groups, including low-income households and ethnic minorities, exhibited poorer sleep routines and thus higher mental disorder risks, underscoring existing social inequalities. Promoting regular sleep behavior may therefore have significant public health benefits.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12889-025-24794-7.

New evidence suggests that air pollution is associated with type 2 diabetes (T2D), but the underlying mechanisms are unclear. Therefore, we aimed to investigate the relationship between air pollution exposure and T2D and to quantify the role of sleep quality in this association. A total of 226,188 participants free of T2D at baseline were enrolled in this prospective cohort study. Annual average concentrations of particulate matter (PM2.5 and PM10) and nitrogen oxides (NO2, NOX) were assessed using a land-use regression. Stratified Cox regression models and Aalen additive hazards models were used to estimate the association between air pollutants and T2D. Causal mediation modelling was used to estimate the indirect impact of air pollutants on T2D through sleep quality. The adjusted hazard ratio (HR) of T2D for a 10-µg/m³ increase in PM2.5, PM10, NO2, and NOX were 1.76 (95%CI: 1.37-2.27), 1.61 (95%CI: 1.36-1.91), 1.05 (95%CI:1.01-1.09), and 1.03 (95%CI:1.02-1.05), respectively. Additionally, for a 10-µg/m³ increase in PM2.5, PM10, NO2 and NOX, estimates of 161 (95%CI: 91-231), 124 (95%CI: 80-168), 14 (95%CI: 4-24) and 10 (95%CI: 5-15) additional T2D cases per 105 persons per year, respectively, were detected. Sleep quality mediated 19.12% to 37.28% of the air pollution-T2D relationship. The study revealed that long-term exposure to air pollution increased the risk of T2D incidence, with sleep quality serving as an important mediating factor. There are two modifiable risk factors for T2D identified and emphasized in our study: air pollution and sleep quality, which shed light on prospective primary prevention.

Abstract Climate change is expected to increase the severity of hurricanes and tropical storms, posing significant risks to the electricity grid. These include downed power lines, damaged solar panels, and impaired wind turbines from high winds. New York (NY) and New Jersey (NJ) are not spared from these vulnerabilities and must strengthen their infrastructure and mitigate social and technical impacts. Clean energy mandates, such as NJ’s Executive Orders No. 315 and 307 (100% clean energy by 2035 and 11 GW of offshore wind by 2040), and NY’s Executive Order No. 166 (40% emissions reduction by 2030), add urgency to ensuring grid resilience under extreme weather. This study demonstrates the power system cyclone impact model (PCIM), used alongside the GenX electricity system planning tool, to assess grid resilience under hurricane-induced high wind speeds in the NY and NJ region. Results reveal that onshore and offshore wind could contribute additional power during storms, provided transmission and storage systems remain operational. This output helps offset outages elsewhere in the grid across all storm categories. In contrast, solar emerges as a vulnerability due to combined impacts from wind stress and cloud cover, significantly reducing generation during and after storms. Thermal generators show the lowest failure rates, though this may partly reflect current model limitations, as only wind stress and cloud cover are considered, excluding hazards like flooding. Non-served energy costs vary with electricity demand and fluctuations in wind and solar output. July stands out as the most vulnerable month, due to high demand and limited wind generation, leading to higher non-served energy. This research provides a first step toward understanding storm-related grid resilience in NJ and NY. The PCIM is designed to be generalizable, with future work focused on expanding its scope to include additional hazards like storm surge and flooding, and more storm-prone regions.

BackgroundAccess to healthcare and socioeconomic deprivation are intricately linked. No studies have been led to measure the effect of healthcare accessibility on mortality in patients with MS so far. The objective was to examine the influence of travel time to the expert MS centre and of the accessibility to primary healthcare services on excess mortality in MS.MethodsA retrospective observational cohort study recruited patients from 18 French MS expert centres, with an onset of MS between 1960 and 2015 and a follow-up of up to 30 years. Primary health facility accessibility was measured by the Spatial aCcessibility multiscAlar index. Specialist care accessibility was measured by road travel time to the expert MS centre. Excess death rates (EDR) and excess hazard ratios were studied using additive excess hazard models with multidimensional penalised splines.ResultsThe study included 33,697 patients. Patients with relapsing-onset MS (R-MS) with a travel time of 40 min had the lowest EDR (Men: 1.2 deaths per 100 person-years (95%CI [0.8;1.8]), women: 0.8 deaths per 100 person-years 95%CI[0.6;1.2]), lower than patients who lived further from the centre. No effect of primary care access was found for patients with R-MS, and no effect of accessibility to primary or specialised care was found for patients with primary progressive MS.ConclusionThis study reveals the impact of travel time to neurologists on excess mortality in patients with R-MS in France. This distance bias association highlights the importance of preventing a potential selection of patients followed in MS expert centres.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12883-025-04459-z.

Smoking and obesity are important modifiable risk factors for ischemic heart disease (IHD), often clustering within the same individuals. Previous US studies showed mixed findings regarding their interaction effects on IHD mortality and only investigated the question on the multiplicative scale, while additive scale is better suited to inform public health interventions. We linked the 1997-2018 National Health Interview Survey data to the 2019 National Death Index. A total of 579503 adults aged 18 years and older were included. Mortality status or last presumed alive was assessed until 31 December 2019. We used Aalen's additive hazards models and calculated the relative excess risk due to interaction (RERI) from Cox proportional hazards models and Fine-Gray subdistribution models that accounted for competing risks to comprehensively evaluate the interaction effect of smoking with obesity on IHD mortality. During 10.4 years of follow-up on average, 13231 IHD deaths occurred. The weighted mortality rate was 177.0 (95% CI: 172.3-181.7) per 100000 person-years (PY). The combination of current everyday smoking and obesity was associated with 55.56 (95% CI: 30.37-80.74) additional deaths per 100000 PY compared to the sum of their individual effects. This additive interaction was supported by multiplicative interactions (HR = 1.19, 1.03-1.39; HR = 1.40, 1.22-1.59) and large RERIs of 1.00 (0.59-1.40) and 0.85 (0.6-1.09) from the Cox and Fine-Gray models, respectively. Our findings highlight the importance of evaluating interactions on multiple scales, which reduces scale-dependence of the interaction effect and can translate better into public health strategies.

High-dimensional error-prone survival data are prevalent in biomedical studies, where numerous clinical or genetic variables are collected for risk assessment. The presence of measurement errors in covariates complicates parameter estimation and variable selection, leading to non-convex optimization challenges. We propose an error-in-variables additive hazards regression model for high-dimensional noisy survival data. By employing the nearest positive semi-definite matrix projection, we develop a fast Lasso approach (semi-definite projection Lasso, SPLasso) and its soft thresholding variant (SPLasso-T), both with theoretical guarantees. Under mild assumptions, we establish model selection consistency, oracle inequalities, and limiting distributions for these methods. Simulation studies and two real data applications demonstrate the methods' superior efficiency in handling high-dimensional data, particularly showcasing remarkable performance in scenarios with missing values, highlighting their robustness and practical utility in complex biomedical settings.

Abstract Extreme weather is increasingly driving human displacement worldwide, a trend expected to worsen with climate change. Quantifying global displacement risk is thus crucial for assessing potential impacts and informing long-term strategies to build more resilient societies, and reducing this risk. One approach involves leveraging classic probabilistic risk modelling methods that hinge on the interplay of hazard, exposure, and vulnerability. Here, we present a methodological stocktaking of these natural-hazard risk models as applied to human displacement. Specifically, we present a globally consistent displacement risk model from multiple hazards under present-day and future conditions. We model population displacement from tropical cyclone winds, coastal floods, river floods, and droughts under present, optimistic, and pessimistic future climate conditions for the middle and end of the century, assuming constant exposure and vulnerability. Our results reveal that current displacement risk is on the order of 30 million annual average displacements (AAD). By 2100, global displacement risk could increase by 75% (157%) under optimistic (pessimistic) climate scenarios. While our risk model makes methodological advances through its global setup, utilisation of two risk frameworks and state-of-the-art datasets, we also highlight current challenges in displacement risk modelling. For instance, our approach primarily models displacement as the direct result of loss of homes from sudden-onset hazards. While we begin to incorporate indirect drivers, such as livelihood loss in river floods and droughts, the model still omits important social, political, and economic dimensions. Nevertheless, as our model adopts a modular design, continuous updates enable the inclusion of additional hazards, improved data, and integration of these broader dimensions. This stocktaking represents a concerted research effort, and our modelling framework may help inform global discussions in international climate negotiations, including those related to Loss and Damage, national action plans, policy development, and other climate adaptation strategies, provided appropriate data and context are applied.

PurposeExposure of atomic bomb (A-bomb) survivors to non-initial (residual) radiation and consequent health effects has not yet been reliably estimated. This study aimed to quantify the contribution of non-initial radiation to the increase in solid cancer mortality risk among A-bomb survivors in Hiroshima through a comparative analysis considering geographical factors.DataWe analyzed the data of 43,056 (17,603 men and 25,453 women) A-bomb survivors registered in the A-bomb Survivor Cohort Database (ABS) at Hiroshima University. These subjects were aged <50 years old at the time of the bombing and lived in Hiroshima Prefecture as of 1 January 1970, after being exposed within 5.0 km of the hypocenter.MethodsThe radiation doses and excess deaths from all solid cancers of the A-bomb survivors were estimated for districts geographically divided by distance and direction from the hypocenter. The dose was defined as the sum of the initial and non-initial radiation doses, and district-averaged non-initial doses were calculated. The excess relative risks (ERRs) of all solid cancer deaths were estimated using multivariate survival analysis with an additive parametric hazard model under the linear no-threshold (LNT) hypothesis. The γ-ray equivalent doses (Sv) from non-initial radiation were estimated based on the estimated ERRs.ResultsEstimated ERRs were notably higher west of the hypocenter than in the other directions. This trend increased with increasing distance from the hypocenter, and the ERRs in men were higher than those in women. Significantly higher ERR values of 52% (p < 0.01) for men and 29% (p < 0.05) for women were obtained at a distance of 2.0–2.5 km west of the hypocenter. The γ-ray equivalent doses estimated from these ERRs exceeded 2 Sv of the effective dose in men west of the hypocenter. This level was notably higher than the estimated initial radiation dose.ConclusionThe findings of this study highlight the considerable contribution of non-initial radiation to the health consequences of the A-bomb survivors. These effects are attributable to the radionuclides generated by the A-bomb detonation, which were assumed to be carried by the wind to the west and deposited with rain in the western region from the hypocenter.

How early-life adversity affects the risk of pediatric-onset immune-mediated inflammatory disease.

BackgroundWaterpipe (WP) use poses not only a risk of nicotine dependence but also additional health hazards. This study examined trends in WP use in Germany, focusing on prevalence by age group and differences in initiation age.MethodsWe analysed data from 76,239 respondents (≥ 14 years) from the German Study on Tobacco Use (DEBRA); a series of bi-monthly national surveys using face-to-face interviews at home (2018 – 2024). Prevalence trends were modelled using binomial logistic regression models with restricted cubic splines.ResultsThe prevalence of WP use decreased over time, to an estimated 0.9 % (95 % CI = 0.6 – 1.2) by mid-2024. This prevalence is made up of 0.1 % (95 % CI = 0.0 – 0.2) 14- to 17-year-olds, 0.3 % (95 % CI = 0.2 – 0.6) 18- to 24-year-olds, 0.3 % (95 % CI = 0.2. – 0.4) 25- to 39-year-olds, and 0.2 % (95 % CI = 0.1 – 0.3) people aged 40 years and older. WP use increased until 2020 up to 2.8 % (95 % CI = 2.3 – 3.4), remained stable for two years and then decreased, especially among people between 25 and 39 years of age. The proportion of 14- to 17-year-old users and users aged at least 40 years remained stable over the years at a low level. Median initiation age was 18 years (25th percentile: 16 years; 75th percentile: 22 years). A lower initiation age was associated with male gender and lower income.ConclusionsWP use increased from 2018 – 2020, stabilised from 2020 – 2022, and then decreased until 2024. Median initiation age was 18, with males and people with lower income starting at a younger age. Targeted public health interventions, focusing on younger males and those with lower socioeconomic status, are needed to prevent early use.

In the United States, the educational gap in all-cause mortality and life expectancy has dramatically increased since 2010. This study investigated whether alcohol use has contributed to the increasing educational gap in mortality by testing the three-way interaction of alcohol use, education and period on all-cause mortality. Cohort study with 9years' follow-up on average. United States. 207 223 males and 255 833 females aged 25 years and older from the 2000-2018 National Health Interview Survey. The outcome was time to all-cause death or last presumed alive by 12/31/2019 based on the National Death Index. Three-way interaction effects between educational attainment (bachelor degree or more vs. high school degree or less), alcohol use (drinking >60 g/day in males and >40 g/day in females vs. lifetime abstinence) and period (after vs. before 2010) were investigated on the multiplicative and additive scales using Cox proportional hazards and Aalen's additive hazards models, respectively, with age as the time scale. Analyses were stratified by sex and adjusted for marital status, race and ethnicity, smoking status, body mass index, physical activity and self-rated health status. During the follow-up period, 30 467 and 34 618 deaths occurred in males and females, respectively, with a pronounced educational gradient. In males, the differential vulnerability to high-level alcohol use by educational attainment substantially increased after 2010 than before 2010 on both multiplicative and additive scales. Specifically, the relative all-cause mortality risk associated with drinking above 60 g per day (vs. lifetime abstinence) in males with low vs. high education increased by 89% after 2010 compared with before 2010 [three-way interaction term: hazard ratio (HR) = 1.89, 95% confidence interval (CI) = 1.03-3.47, P = 0.04 from the Cox model]. This result was further supported by the Aalen's model, indicating that the educational difference in mortality risk linked to drinking above 60 g per day increased by 8.85 additional deaths per 1000 person-years (95% CI = 0.90-16.79, P = 0.029) after 2010. No change of differential vulnerability to alcohol use was found in females. Alcohol use appears to be a key element in the widening educational gap in all-cause mortality risk in males after 2010 in the United States.

Clostridioides difficile is a leading cause of health care-associated infections. Understanding the association among C difficile carriage, antibiotic use, and infection hazard is essential for infection prevention. To evaluate the hazard of C difficile infection (CDI) among asymptomatic carriers vs noncarriers of C difficile and whether it is associated with antibiotic exposure. This retrospective cohort study conducted between June 18, 2017, and June 21, 2023, analyzed hospitalizations from Sheba Medical Center in Ramat Gan, Israel, which routinely screens for C difficile in high-risk patients admitted to internal medicine. Adult patients (aged >18 years) without active CDI at admission were included. Antibiotic exposure during hospitalization, including specific classes. The primary outcome was the development of CDI, as confirmed by laboratory testing for C difficile. Antibiotic exposure was assessed as a time-varying variable. The study included 33 756 hospitalizations among 23 001 patients (median [IQR] age, 78 [68-87] years; 52.8% men). C difficile infection occurred in 67 of 1624 hospitalizations (4.1%) with positive screening results and in 47 of 32 132 hospitalizations (0.1%) with negative screening results. A positive C difficile screening result at admission was associated with a high hazard of infection (hazard ratio [HR], 27.5; 95% CI, 18.7-40.3). Antibiotic exposure was associated with an increased hazard for CDI (HR, 1.98; 95% CI, 1.24-3.16). Piperacillin and tazobactam showed the most pronounced hazard for CDI (HR, 2.18; 95% CI, 1.41-3.36). Among asymptomatic carriers, antibiotic exposure was not significantly associated with a further increase in CDI hazard (HR, 1.07; 95% CI, 0.73-1.58). In this cohort study, carriers of C difficile had a substantially higher baseline hazard for hospital-onset CDI. Antibiotic exposure was associated with an increased hazard among noncarriers but was not significantly associated with additional hazard among carriers. These findings suggest that while antibiotic stewardship may reduce CDI risk in noncarriers, additional strategies may be needed for carriers given their elevated baseline risk.

Abstract Aim The performance of plasma soluble urokinase plasminogen activator receptor (suPAR) for prediction of heart failure (HF) readmission or death within 5 years was assessed in patients incurring (i) initially undifferentiated chest pain and (ii) immediately after diagnosed acute coronary syndromes (ACS), and (iii) in recovery after ACS. Methods suPAR concentrations were measured at admission for acute chest pain patients (n=917) including confirmed ACS (26.5%), and in an independent 4-6 weeks post-ACS cohort (n=1301). suPAR’s prognostic performance, in comparison, and combination with cTnI or NT-proBNP was evaluated by risk discrimination, hazard ratios (HR) as continuous variables and cut-off concentrations across the three settings in unadjusted and adjusted analyses. Results In acute undifferentiated chest pain including the subgroup with confirmed ACS and separately post-ACS convalescence, combining suPAR and NT-proBNP yielded the highest discriminatory power for endpoints (c-statistics &amp;gt;0.80, ≥ Δ0.02). suPAR in the acute and post-ACS convalescent conferred additional hazard for the composite endpoint of HF/death (HR&amp;gt;1.4), HF (HR&amp;gt;1.3) and death (HR&amp;gt;1.2) after adjustment for risk factors including circulating cardiac markers. Although suPAR &amp;gt;3.65ng/mL (&amp;gt;83% specificity, &amp;gt;91% NPV) was superior in admission ACS (HR:10.5) for risk of HF/death, independent prediction for the composite endpoint remained consistent across the three settings and in men. Conclusion Plasma suPAR independently predicted risk of readmissions with HF and death, independent of key clinical indicators and cardiac markers at all points of the patient journey from acute chest pain presentation through to post-ACS convalescence. Its use in acute chest pain and ACS may augment risk stratification strategies.

Introduction: Variable selection is increasingly becoming a key step in biomedical research, particularly in high-throughput genomic data analysis. One major focus is selecting relevant gene expression profiles associated with time-to-event outcomes, such as death. A significant challenge in this context is competing risks, where identifying a small subset of gene expression profiles related to the cumulative incidence function (CIF) is essential. Methods: Several methods have been proposed for directly modeling CIF, primarily by modeling the subdistribution hazard function for the event of interest. We proposed a regularized method for variable selection in the additive subdistribution hazards model by integrating five penalized likelihood approaches—Least Absolute Shrinkage and Selection Operator (LASSO), Adaptive LASSO (ALASSO), Elastic Net (ENET), Adaptive Elastic Net (AENET), and Smoothly Clipped Absolute Deviation (SCAD)—with the pseudoscore method. We conducted Monte Carlo simulations to evaluate the performance of our proposed method. Furthermore, the method was applied to a publicly available dataset of 301 patients diagnosed with non-muscle-invasive bladder carcinoma from five countries between 1987 and 2000. Results: Our proposed method was evaluated through simulation studies and applied to genomic data, focusing on progression-free survival as the endpoint and identifying the genes associated with the CIF of bladder cancer in the presence of competing events. Five genes, namely DCTD, IGF1R, NCF2, PLEK, and CDC20, were consistently identified across different penalties. Notably, the overexpression of DCTD and IGF1R was associated with a decreased cumulative incidence of bladder cancer progression or death. In contrast, the overexpression of NCF2, PLEK, and CDC20 correlated with an increased cumulative incidence of these events. Conclusion: According to the findings of the simulation studies, all penalties yielded comparable results in terms of sensitivity and specificity. However, the AENET and ALASSO ppenalties demonstrated superior estimation accuracy.

Occurrence, dermal and unintentional ingestion exposure, and human health risk of additives and chemical allergens in commercial mouse pads.