Bayesian Hierarchical Model Research Articles

Using hierarchical drift diffusion models to elucidate computational mechanisms of reduced reward sensitivity in adolescent major depressive disorder

BackgroundAnhedonia—a core symptom of major depressive disorder (MDD)—is closely related to diminished reward sensitivity. Nonetheless, the psychopathological and computational mechanism underlying anhedonia in young patients with MDD remains unclear. Therefore, this study aims to investigate reward sensitivity in adolescents and young adults with MDD using computational modelling.MethodsOverall, 70 patients with MDD and 54 age- and sex-matched healthy controls (HC) completed a probabilistic reward task (PRT) to assess their general behavioral inclination towards more frequently reinforced stimuli (i.e., “response bias”). Bayesian hierarchical drift diffusion modeling (HDDM) was employed to determine changes in reward sensitivity and computational process during decision-making.ResultsAdolescents with depression showed a trend toward reduced response bias compared to those in HC. HDDM analysis revealed wider decision thresholds in both adolescents and young adults with MDD group. Adolescents with MDD exhibited significantly lower drift rates and reduced starting point bias compared to those in HC. Higher anhedonia levels were linked to lower drift rates and wider decision thresholds. Additionally, increased discriminability correlated with higher drift rates, while higher response bias was linked to larger starting points.ConclusionsOur findings suggest that reduced reward sensitivity and slower evidence accumulation during reward learning may serve as potential indicators of anhedonia in adolescents with MDD. These findings provided crucial insights into the dysregulated positive affect model, underscoring a dysfunctional reward system as a key factor in anhedonia developmental psychopathology in depression.

Relational Resonance and Content Creation

This study investigates how tie formation shapes the production of user-generated content on social media, offering insights into enhancing user engagement and alleviating social media fatigue. We conceptualize tie formation as a relationship codevelopment process involving a tie initiator and a respondent. Drawing on social psychology literature, we propose a framework that traces how initiators’ social cognition evolves throughout this process—from proactively extending an inviting social tie to receiving a responsive one in return. Central to our framework is a perceptual mechanism we term relational resonance, which fosters intrinsic motivation and prompts initiators to adjust their content creation. Using data from YouTube, we develop a Bayesian hierarchical model to identify individual creators’ content adjustments, measured by changes in video volume and viewer engagement. Our results reveal that while initiators tend to reduce content production after receiving a responsive tie, the videos they create generate higher engagement, with “like” and “subscribe” rates increasing by 1.1 and 2.9 percentage points, respectively. Remarkably, the receipt of a responsive tie leads to average weekly increases of 66 likes and 183 subscriptions even after accounting for the decrease in video volume, reinforcing the “quality over quantity” principle in content creation strategies. An online experiment further elucidates the mechanism underlying the observed adjustment behavior, providing robust validation for our proposed framework. Implications for research and practice are discussed. This paper was accepted by D.J. Wu, information systems. Funding: T. Song was supported by National Natural Science Foundation of China [Grants 71902114 and 72372101]. Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2023.00830 .

Bayesian hierarchical model for heterogeneity in the diagnosis and treatment of hepatocellular carcinoma (HCC) in the African continent

HCC prevalence and etiology differ between African regions. To better understand these differences, a survey was conducted to characterize the health care practices in HCC management in Africa. Due to limited number of respondents from certain countries in Africa and due to scarce availability of resources, the traditional method of adaptive quadrature produces unstable estimates in this complex setting with sparse multinomial outcomes and with sparse clustered random effects. We propose a Bayesian hierarchical model to address this challenge, apply it to the sparse real data captured in the survey, and validate the proposed method through simulations on synthetic data.

Exploring safety effects on urban expressway diverging areas: crash risk estimation considering extreme conflict types

Previous research solely employed a single type of conflict extremes for crash estimation, without considering the joint impact of multiple types of conflict extremes on crash risk. Therefore, two analysis frameworks based on conflict extremes were proposed: separate modeling and cooperative modeling. Based on the trajectories from five diverging areas, longitudinal and lateral conflicts were extracted. Then, a Bayesian hierarchical model for joint multi-location conflict extremes was constructed. Next, the threshold for conflict extremes was determined using automatic mean residual life plots, and a link function was established between the logarithmic scale parameter and dynamic and static variables. Finally, model parameters were estimated using the Markov Chain Monte Carlo simulation method, and a comparative analysis of crash probabilities and overall risks for diverging areas in the two frameworks was conducted by the fitted distributions. The results show that density differences, speed differences, and the ratio of large vehicles are important covariates explaining the non-stationarity of conflict extremes. In terms of crash probability, significant covariates exhibit stronger explanatory power for longitudinal conflicts compared to lateral conflicts. At the overall risk level, the accuracy of the separate modeling is higher compared to the cooperative modeling.

Unveiling pandemic patterns: a detailed analysis of transmission and severity parameters across four COVID-19 waves in Bogotá, Colombia

BackgroundDespite a wealth of data from high-income countries, there is limited information on the distinct epidemiological patterns observed in diverse, densely populated regions within Latin America. This retrospective analysis of COVID-19’s four major waves in Bogotá, Colombia, evaluates 1.77 million cases in detail.MethodsA comprehensive suite of statistical methods was employed. Transmission dynamics were assessed by estimating the instantaneous reproduction number R(t), while variant-specific transmission advantages were estimated using multinomial logistic regression models. Disease severity was assessed through a suite of indicators: Hospitalisation Case Ratio (HCR), intensive care unit case ratio (ICU-CR), case fatality ratio (CFR), hospitalisation fatality ratio (HFR), and ICU fatality ratio (ICU-FR). Additionally, we analysed the distribution of hospitalisations, ICU admissions, and fatalities by age group and wave. We employed a Bayesian hierarchical model to capture epidemiological delays—such as onset-to-death, hospitalisation, and ICU admission durations to estimate hospital and ICU stay durations.ResultsOur findings reveal substantial variation in R(t), with peaks exceeding 2.5 during the ancestral and Omicron waves. Over the course of the pandemic, we observed a 78% reduction in CFR, underscoring shifts in clinical severity. The third wave, associated with the Mu variant, recorded the highest case and death counts, alongside a decreased CFR, an elevated HFR, and a shift in the most affected age group towards younger populations. In contrast, the fourth wave, driven by the Omicron variant, exhibited the highest reproduction number and the lowest overall severity. This wave was characterised by a significant increase in pediatric hospitalisations. The study reveals a continued decline in the mean durations of hospital and ICU stays across the four waves, with hospital stays decreasing from 10.84 to 7.85 days and ICU stays dropping from 16.2 to 12.4 days.ConclusionsThis study reveals significant shifts in transmission and severity metrics—including mortality, hospitalisation, and ICU rates and stays—across age groups during Bogotá’s four COVID-19 waves. These insights underscore the value of retrospective analyses to understand the pandemic’s varied impact and inform public health strategies in diverse urban settings.

Projecting future migration with Bayesian hierarchical gravity models of migration: an application to Africa

In this paper, I present and discuss a novel approach to parameterize a gravity model of migration using Bayesian hierarchical models with random intercepts that are free to vary by country of origin, destination, and directed origin-destination country pairs. I then utilize this model to project transboundary migration flows between African countries to the horizon 2050. To do so, I use data on projected future crop yields and water availability from the ISIMIP2b scenarios in combination with projections on future economic and demographic trends from the Shared Socio-Economic Pathways (SSPs). The results indicate that over the period 2010–2050 between 8 to 17 millions people are projected to migrate internationally on the African continent. Yet, only a small portion of these migrants will be induced to move because of climate change. To the contrary, comparisons between SSPs scenarios suggests that economic development will have a far larger impact on projected level of international migration on the continent than climate change.

The longitudinal association between reward processing and symptoms of video game addiction in the Adolescent Brain Cognitive Development Study.

Video games are a common form of entertainment in adolescents, which may result in gaming habits characterized by impairment to reward-related decision-making. The aim of the current study was to investigate the relationship between reward processing and symptoms of gaming addiction in adolescents. Data from three consecutive follow-up years (years 2, 3 and 4) of theAdolescent Brain Cognitive Development (ABCD) Study were analyzed (n = 6,143, total observations= 12,745, mean age at year-2 = 12 years). Participants completed the Video Game Addiction Questionnaire (VGAQ) at each visit. Discrete stages of reward processing were measured at the year-2 visit using the Monetary Incentive Delay task while the participant completed a functional magnetic resonance imaging (fMRI) scan. Bayesian hierarchical linear models were employed to examine the longitudinal association between reward processing in regions of interest at year-2 and VGAQ scores over time. Lower activation in the bilateral caudate during the anticipation of a large reward (β = -0.87, 95% CI: -1.68, -0.07) was associated with greater VGAQ scores over time. This implies that for each one-unit increase in brain activity in the caudate, there was an associated 0.87-point decrease in symptoms of gaming addiction as measured by the VGAQ. No association was found between reward feedback and VGAQ scores. The findings suggest that abnormal reward processing in the caudate nucleus is associated with symptoms of gaming addiction in adolescents. These results provide a clearer understanding of the brain mechanisms involved in gaming addiction, which could inform future preventive and therapeutic strategies.

The blinding status and characteristics in acupuncture clinical trials: a systematic reviews and meta-analysis

BackgroundSham acupuncture is a widely accepted control in acupuncture clinical trials. Given the nature of acupuncture, it is warranted to assess the blinding of sham-controlled trials. Despite the sham acupuncture design having been widely used, the overall blinding of sham acupuncture and the characteristics of blinding assessment in acupuncture trials are unclear. This research aims to assess the blinding status of acupuncture clinical trials and explore the blinding assessment characteristics in acupuncture trials.MethodsThis meta-analysis included all the acupuncture clinical trials published in English that performed blinding assessments and reported the results. We searched PubMed, Embase, and Web of Science for randomized controlled trials (RCTs) from inception to April 2024. The primary outcome is Bang’s Blinding Index (Bang’s BI) and 95% credibility interval (CrI) was pooled using a Bayesian hierarchical model. The study adheres to the PRISMA guidelines.ResultsSixty-four eligible studies published from 1999 to 2024 were included. The mean of Bang’s BI was − 0.24 (95% CrI − 0.34 to − 0.14, tau2 = 0.13) for the sham acupuncture group and 0.41 (95% CrI 0.32 to 0.49, tau2 = 0.10) for the verum acupuncture group. The characteristics of blinding showed that 62.50% of the trials had a Bang’s BI greater than 0 in the verum group and less than 0 in the sham group; in 28.15% of the trials, the Bang’s BI was greater than 0 in the verum group and greater than 0 in the sham group. Subgroup analysis revealed that area, number of research centers, treatment sessions, acupoints number, and evaluation timepoint can influence blinding results.ConclusionOverall blinding status in current acupuncture clinical trials shows a majority correctly guessing for the verum group and opposite guessing for the sham group. However, in some acupuncture trials, the blinding of sham acupuncture might be compromised. Factors such as the Asian population, penetrating sham needling, and querying participants about their group assignment during the study increase the risk of unblinding and warrant careful consideration in sham acupuncture control design. Furthermore, researchers should closely monitor the blinding status of sham acupuncture and transparently report details of blinding assessments.Systematic review registrationPROSPERO CRD42023403595.

Optimal yoga dose for improving cancer-related fatigue, quality of life, and sleep quality in breast cancer survivors: a Bayesian dose-response meta-analysis.

During long-term treatment, breast cancer survivors (BCSs) often experience decreased sleep quality and quality of life and increased fatigue. The interplay between these factors continually worsens the physical and psychological states of BCSs. Although yoga exercise has been shown to improve these adverse outcomes, the exact dose required is unknown. The purpose of our study was to determine the optimal yoga dose for improving cancer-related fatigue, sleep quality, and quality of life in BCSs. A systematic search of MEDLINE, Embase, the Cochrane Library, and Web of Science was conducted from the beginning of the database search until March 2024 to identify randomized controlled trials on the effects of yoga exercise on BCS-specific outcomes. Paired and dose-response meta-analyses were performed using Bayesian hierarchical random-effects models to analyze these effects. A total of 2423 breast cancer survivors were included in the 32 studies, with 1252 in the yoga intervention group and 1171 in the control group. Yoga intervention improved fatigue (standard mean difference [SMD]: - 0.55; 95% CrI, - 0.64 to - 0.45. Standard deviation [SD]: 0.32; 95% CrI, 0.23 to 0.45. RSRF ≤ 1.01), sleep quality (SMD: - 0.29; 95% CrI, - 0.43 to - 0.15. SD: 0.56; 95% CrI, 0.35 to 0.88. RSRF ≤ 1.01), and quality of life (SMD: 0.42; 95% CrI, 0.30 to - 0.53. SD: 0.32; 95% CrI, 0.23 to 0.45. RSRF ≤ 1.01) in BCS. It has a nonlinear dose-response relationship, a U-shaped dose-response relationship with fatigue and quality of life, and a positively correlated dose-response relationship with sleep quality. The optimal exercise dose for improving both fatigue and quality of life was 510 METs-min/week, and sleep quality significantly improved when the dose exceeded 160 METs-min/week. Yoga was less effective for fatigue during treatment compared to post-treatment, while its effects on quality of life and sleep quality remained relatively stable throughout both periods. Furthermore, with increasing age, the benefits of yoga interventions for improving fatigue and sleep quality in BCS decrease, while the benefits for improving quality of life increase. This meta-analysis provided critical information on the optimal weekly dose of yoga exercise (220min/week) to improve fatigue, sleep quality, and quality of life in patients with BCSs. Practitioners should consider the therapeutic stage and age of BCSs when developing yoga strategies to maximize the benefits. For BCS, personalized yoga interventions tailored to treatment stages and age, coupled with regular effectivenessassessments, can significantly enhance recovery and improve quality of life.

Inferring failure risk of on-site wastewater systems from physical and social factors

Aging infrastructure and climate change present emerging challenges for clean water supply and reliable wastewater services for communities in the United States (US). In Georgia, for example, the failure rates of on-site wastewater systems (OWTS) have increased from 10% to 35% in the last two decades as the systems age. In this work, we develop a hierarchical Bayesian model to understand the different contributions of physical and social factors driving OWTS failures using a long-term collection of 201,000 Georgia’s OWTS inspection records. The out-of-sample validation accuracy of our hierarchical Bayesian model is 70% within Georgia, outperforming other machine learning models that do not consider the multiscale nature of the problem. Overall, we find counties that experience more extreme precipitation and are situated in steeper-sloped regions are significantly associated with increased failure risks. Uncertainties, meanwhile, are largely associated with counties experiencing more precipitation and have lower median housing value.

Hotspot mapping of pest introductions in the EU: A regional analysis of environmental, anthropogenic and spatial effects

Plant pests may pose a significant threat to global agriculture, natural ecosystems and biodiversity, causing severe ecological and economic damage. Identifying regions more susceptible to pest introductions is crucial for developing effective prevention, early detection and outbreak response strategies. While historical data on pest introductions in the European Union (EU) exist, they are typically reported at the regional level. This broad aggregation has posed a challenge for accurate analysis in plant health research. This study addresses this gap by leveraging existing regional data to identify hotspots of pest introductions within the EU and the UK, through a Bayesian hierarchical spatial model. Specifically, we employed the Besag, York, and Mollié (BYM) model to identify higher risk regions (NUTS2) by incorporating covariates and spatial effects to consider information from neighbouring areas. The results showed a positive effect of annual average temperature, annual average precipitation, and human population density on the risk of pest introduction, highlighting the relevance of the spatial component. Our analysis pinpoints high-risk regions in southern Europe, particularly northern Italy. Additionally, the high human population density and documented pest introductions in the Netherlands contributed to its elevated risk. While limitations exist due to the regional nature of the data, this study represents a methodological advancement, demonstrating the effectiveness of spatial models and offering a robust framework for future studies using regional data. It also provides insights that can inform targeted prevention, early detection and preparedness strategies, ultimately contributing to safeguarding agriculture, natural ecosystems and biodiversity in the EU and UK.

Hierarchical Bayesian modeling for Inverse Uncertainty Quantification of system thermal-hydraulics code using critical flow experimental data

The best estimate plus uncertainty methodology in nuclear system thermal-hydraulic studies necessitates a comprehensive understanding of uncertainties in system code predictions. The forward uncertainty quantification (UQ) process involves the propagation of input uncertainties through the computational models to obtain uncertainties in the outputs. To this end, achieving an accurate estimation of input uncertainties is important, which is the focus of inverse UQ (IUQ). Traditionally, research in Bayesian IUQ within the nuclear engineering domain has largely relied on single-level Bayesian inference. While being effective for relatively small datasets, this approach encounters limitations for cases with large datasets. The use of a single-level model may prove inefficient, as the resultant posterior distributions can significantly differ when distinct subsets of data are employed. To address this issue, we employ an hierarchical Bayesian model for IUQ. This approach involves organizing observations into different groups based on the test conditions, thereby accommodating varying calibration parameters across these distinct groups. In this study, we developed and implemented a hierarchical Bayesian IUQ method to consider the grouping effect of critical flow measurement data from various geometries. Comparing the outcomes of IUQ under different selections of test data using hierarchical Bayesian IUQ against those obtained from single-level Bayesian IUQ, the forward propagation of hierarchical Bayesian IUQ results demonstrates a notably improved agreement with the experimental data.

Spatial and Temporal Bayesian Hierarchical Model Over Large Domains With Application to Holocene Sea Surface Temperature Reconstruction in the Equatorial Pacific

AbstractWe present a novel space‐time Bayesian hierarchical model (BHM) to reconstruct annual Sea Surface Temperature (SST) over a large domain based on SST at limited proxy (i.e., sediment core) locations. The model is tested in the equatorial Pacific. The BHM leverages Principal Component Analysis to identify dominant space‐time modes of contemporary variability of the SST field at the proxy locations and employs these modes in a Gaussian process framework to estimate SSTs across the entire domain. The BHM allows us to model the mean field and covariance, varying in space and time in the process layers of the hierarchy. Using the Markov Chain Monte Carlo (MCMC) method and suitable priors on the model parameters, posterior distributions of the model parameters and, consequently, posterior distributions of the SST fields and the attendant uncertainties are obtained for any desired year. The BHM is calibrated and validated in the contemporary period (1854–2014) and subsequently applied to reconstruct SST fields during the Holocene (0–10 ka). Results are consistent with prior inferences of La Niña‐like conditions during the Holocene. This modeling framework opens exciting prospects for modeling and reconstruction of other fields, such as precipitation, drought indices, and vegetation.

Tropical Cyclone Wind Field Reconstruction for Hazard Estimation via Bayesian Hierarchical Modeling With Neural Network

AbstractTropical cyclones (TCs) are one of the biggest threats to life and property around the world. Accurate estimation of TC wind hazard requires estimation of catastrophic TCs having a very long return period spanning up to thousands of years. Since reliable TC data are available only for recently decades, stochastic modeling and simulation turned out to be an effective approach to achieve more stable hazard estimates. In common practice, hundreds of thousands of synthetic TCs are generated first, then wind fields are reconstructed along synthetic TC tracks for hazard estimation. A Bayesian hierarchical modeling approach to the reconstruction of TC wind field is proposed. A modified Rankine vortex is adopted as the wind field model, of which the four free parameters are modeled simultaneously through a multi‐output neural network as a latent process of the wind field. The four parameters are finally represented, spatially and temporally, by a set of neural network weights, The Bayesian model averaging technique is used for parameter estimation and wind field reconstruction, based on a ensemble of maximum a posteriori estimates of the set of weights. Together with previously proposed algorithm for synthetic TC simulation, a two‐stage scheme for TC wind hazard estimation has been formed, which is based on best‐track data only and thus is highly consistent. Application of this scheme to the offshore waters in the western North Pacific basin shows inspiring performance and great flexibility for various purposes of TC wind hazard estimation.

Effects of concurrent aerobic and strength training in patients with type 2 diabetes: Bayesian pairwise and dose-response meta-analysis

This study aimed to investigate the effects of concurrent aerobic and strength training (CT) in patients with type 2 diabetes and determine the most effective dose of CT. From the...

Kalkayotl 2.0. Bayesian phase-space modelling of star-forming regions, stellar associations, and open clusters

Star-forming regions, stellar associations, and open clusters are fundamental stellar systems where predictions from star-formation theories can be robustly contrasted with observations. We aim to provide the astrophysical community with a free and open-source code to infer the phase-space (i.e. positions and velocities) parameters of stellar systems with lesssim 1000 stars based on Gaia astrometry and possibly observed radial velocities. We upgrade an existing Bayesian hierarchical model and extend it to model 3D (positions) and 6D (positions and velocities) stellar coordinates and system parameters with a flexible variety of statistical models, including a linear velocity field. This velocity field allows for the inference of internal kinematics, including expansion, contraction, and rotation. We extensively validated our statistical models using realistic simulations that mimic the properties of the Gaia Data Release 3. We applied Kalkayotl to beta -Pictoris, the Hyades, and Praesepe, recovering parameter values compatible with those from the literature. In particular, we found an expansion age of $19.1 Myr for beta -Pictoris and rotational signal of $32\! $ for the Hyades and that Praesepe's rotation reported in the literature comes from its periphery. The robust and flexible Bayesian hierarchical model that we make publicly available here represents a step forward in the statistical modelling of stellar systems. The products it delivers, such as expansion, contraction, rotation, and velocity dispersions, can be directly contrasted with predictions from star-formation theories.

Heart Failure and All-Cause Hospitalizations in Patients With Heart Failure

Heart failure (HF) hospitalization is a common end point in HF trials; however, how HF hospitalization is associated with all-cause hospitalization in terms of proportionality, correlation of treatment effects, and concomitant reporting has not been studied. To determine the ratio of HF to all-cause hospitalizations, whether reported treatment effects on HF hospitalization are associated with treatment effects on all-cause hospitalization, and how often all-cause hospitalization is reported alongside HF hospitalization. PubMed was searched from inception to September 2, 2024, for randomized clinical trials (RCTs) of HF treatments using MeSH (medical subject heading) terms and keywords associated with heart failure, ventricular failure, ventricular dysfunction, and cardiac failure, as well as the names of specific journals. RCTs of HF treatments and reporting on HF hospitalization published in 1 of 3 leading medical journals (New England Journal of Medicine, The Lancet, or JAMA). The PRISMA guidelines were followed. Data extraction was performed by 2 reviewers, and disagreements were resolved by consensus. Trial baseline characteristics and outcome data on HF and all-cause hospitalizations were extracted. The ratio of HF to all-cause hospitalizations was calculated. The association of HF hospitalization effects with all-cause hospitalization effects was evaluated using hierarchical bayesian models with weak priors. The posterior distribution was used to calculate the HF hospitalization treatment effects that would need to be observed before a high probability (97.5%) of a reduction in all-cause hospitalization could be achieved. The proportion of trials reporting all-cause hospitalization was calculated. HF and all-cause hospitalizations. Of 113 trials enrolling 261 068 patients (median proportion of female participants, 25.4% [IQR, 21.3%-34.2%]; median age, 66.2 [IQR, 62.8-70.0] years), 60 (53.1%) reported on all-cause hospitalization. The weighted median ratio of HF to all-cause hospitalizations was 45.9% (IQR, 30.7%-51.7%). This ratio was higher in trials with greater proportions of New York Heart Association class III or IV HF, with lower left ventricular ejection fractions, investigating nonpharmaceutical interventions, and that restricted recruitment to patients with HF and reduced ejection fraction. Reported effects on HF and all-cause hospitalizations were well-correlated (R2 = 90.1%; 95% credible interval, 62.3%-99.8%). In a large trial, the intervention would have to decrease the odds of HF hospitalization by 16% to ensure any reduction, 36% to ensure a 10% reduction, and 56% to ensure a 20% reduction in the odds of all-cause hospitalization with 97.5% probability. In this meta-analysis of HF trials, all-cause hospitalization was underreported despite a large burden of non-HF hospitalizations. Large reductions in HF hospitalization must be observed before clinically relevant reductions in all-cause hospitalization can be inferred.

Blue and fin whale residence time and occupancy in Navy training and testing areas off the U.S. West Coast

Blue (Balaenoptera musculus) and fin (B. physalus) whales are prominent species in the California Current Ecosystem (CCE) using the area for migration and foraging. Blue whales migrate between tropical wintering grounds and higher latitude foraging grounds in the CCE, whereas fin whales may remain there year-round. We used Argos tracks of 182 blue (mean duration: 78 days) and 98 fin (mean duration: 35 days) whales tagged from 1994-2018 along the U.S. West Coast to examine residence time and occupancy in U.S. Navy training and testing areas. Bayesian hierarchical switching state-space model locations calculated from Argos tracks were used to provide residence time and proportion of tracking duration within each Navy area. We also calculated relative occupancy isopleths from the state space models, for both directed and non-directed movement, to assess spatial use of Navy areas. We then used generalized linear mixed models and beta regression to examine relationships between Navy area use and covariates, including six climate indices. Point Mugu Sea Range (PMSR) was the most heavily used Navy area by both species in terms of whale numbers, followed by the Southern California Range Complex (SOCAL). Residence time was longest in PMSR for blue whales and in the Northwest Training and Testing Study Area (NWTT) for fin whales. Blue whale use in PMSR was greater in summer versus autumn and for whales tagged in Southern versus Central California. Proportion of tracking duration decreased in all areas with increasing values of the Oceanic Niño Index (ONI), and in PMSR, with increasing values of the North Pacific Gyre Oscillation. Fin whale residence time was longer in PMSR during summer than winter or spring and decreased with increasing values of Habitat Compression Index and ONI. Proportion of tracking duration for fin whales was greatest in summer in PMSR and in winter in SOCAL and NWTT. Overlap of directed and non-directed movement was greatest in PMSR for both species. This assessment of blue and fin whale residence time and occupancy in Navy areas improves our understanding of potential threats these animals face, not only from military activities in Navy areas, but throughout the CCE.

Pesticides and risk of pancreatic adenocarcinoma in France: a nationwide spatiotemporal ecological study between 2011 and 2021.

While pancreatic adenocarcinoma (PA) incidence is increasing, especially in France, the association between pesticides and PA remains unclear. The aims of this study were to assess the spatiotemporal distribution of the incidence of PA in France between 2011 and 2021 and to determine whether pesticide exposure was associated with higher risk of PA. We employed a disease-mapping and ecological regression approach with medicoadministrative data covering 99% of the French population. Exposure data were drawn from an open purchase database. A pesticide exposure intensity index (PEXI), defined as the logged and scaled median of the ratio quantity of substance over agricultural surface per spatial unit was used for total quantity (total PEXI) and 9 specific substances. The analyses were adjusted for tobacco-induced diseases, deprivation, community medicine accessibility, alcohol-related disease and morbid obesity. A Bayesian hierarchical spatiotemporal model was used to both model the incidence over time and space, and to estimate the risk of PA for pesticide use. We identified 134102 incident cases of PA between 2011 and 2021. The relative risk of PA was heterogeneous across space with greater risk around Paris, central France and the Mediterranean coast. We observed an association, albeit small, between the total PEXI and PA incidence over the study period (RR: 1.0130; CI95% [1.0057;1.0204]). Sulphur for spraying, mancozeb, and glyphosate showedevidence of an association of the same magnitude. These findings show that new cases of PA occur heterogeneously in space, raising questions about our understanding of PA environmental risk factors. The association with pesticide exposure should be confirmed and underlying mechanisms understood using individual-level studies.

Yes, they're all individuals: Hierarchical models for repeat survey data improve estimates of tree growth and size

Abstract Repeat measurement surveys of tree size are used in forests to estimate growth behaviour, biomass and population dynamics. Although size is measured with error and individuals vary in their growth trajectories, current size‐based growth modelling approaches do not usually or fully account for both of these features, and therefore under‐utilise available data. We present a new method that leverages the auto‐correlation structure of repeat surveys into a hierarchical Bayesian longitudinal growth model. This new structure allows users to correct for measurement error and capture individual‐level variation in growth trajectories and parameters. To demonstrate the new method we applied it to a sample of tropical tree survey data from long‐term monitoring sites at Barro Colorado Island. We were able to reduce estimated error in size and growth, and extract individual‐ and population‐level growth parameter estimates. We used simulation to evaluate the ability of the new method to improve estimates of growth rate and size, and estimate individual and species‐level parameters. Our method substantially improved the root mean squared error (RMSE) for growth by an average of 61% compared to existing approaches using pairwise differences and reduced RMSE in estimated size RMSE compared to ‘observed’ values in simulated data. Better numerical integration methods (Runge–Kutta fourth order in comparison to Euler and midpoint) provided better estimates of parameters, but did not improve the estimation of size and growth. The choice of a positive growth function eliminated all negative increments without data exclusion. Overall, this study shows how we can gain new and improved insights on growth, using repeat forest surveys. Our new method offers improved biomass dynamics estimation through reduced error in sizes over time, coupled with novel information about within‐species variation in growth behaviour that is inaccessible with species average models, such as individual parameters for the growth function which allows for relationships between parameters to be considered for the first time.