Generalized Additive Mixed Models Research Articles

What drives wild boar density and population growth in Mediterranean environments?

Accurate prediction of fluctuations of wildlife local number of individuals is crucial for effective population management to minimise human-wildlife conflicts. Climate, habitat, food availability, and density dependence are among the main factors influencing mammalian population dynamics. In southern Europe, precipitation and temperature, particularly during summer have been suggested as key factors affecting wild boar (Sus scrofa L.). However, there is uncertainty regarding the role of these factors and the mechanisms driving population fluctuations. This study utilized long-term data of wild boar populations from 14 study sites collected for 23 years in Catalonia, Spain, to analyse the factors that drive population density and growth rate. Generalized Additive Mixed Models (GAMM) explained respectively, 94 % and 65 % of the density and growth rate variability. Spring precipitation in both current and previous year, female weight, and forest cover (particularly above 60 %) were directly associated with higher wild boar densities and population growth rates. The interaction between crop cover and total annual precipitation also played a significant role in determining population density. Higher densities were linked to lower population growth in the following year, likely due to a density-dependent process. These results suggest that the expected decrease in rainfall linked with global warming may limit the availability of natural resources and potentially slow wild boar population growth. Nevertheless, wild boar can exploit alternative anthropogenic food sources, potentially leading to an increase of human-wildlife conflicts. Therefore, incorporating management policies aimed at restricting wild boar access to human food sources is key for controlling their reproductive output. Additionally, landscape management strategies targeted at diminishing refuge and resource availability in regions experiencing high wild boar impact are essential for contributing to sustainable coexistence between wild boars and human populations.

The Kids are not alright - When did we start getting more more distressed?

IntroductionMuch has been made of the decline in population mental health over COVID but most studies show this just exacerabted a loing term trend This has predominnatly been attributed to changes in adolescent mental health over the past decade but there ahs been little evalaution of whether this post Millenium cohort was the first to demonstrate such a declineObjectivesThis study investigates to what extent mental differs in people born in different decades – i.e., possible birth cohort differences in the mental health of the popualtion over the past two decades To remove the linear dependency and identify any differences in trends between cohorts, we model mental health for each cohort as a nonlinear smooth function of age in an age-cohort model.MethodsThis analysis draws on 20 annual waves of the Household Income and Labour Dynamic in Australia (HILDA) survey.,is a nationally representative household panel that commenced in 2001 with 13,969 participants. The birth cohort of each person was defined by the decade of birth year(1940s, 1950s, etc). Mental ill health was assessed with the MHI5 from the SF36, in each wave and K10 from alternate waves. We estimate and compare penalized smooth trends in mental health for each cohort using restricted maximum likelihood (REML) using generalized additive mixed modelling (GAMM). Cohort effects are captured by directly estimating the differences between the smooth age trends of adjacent cohorts.ResultsLater cohorts were more likely to have poorer mental health, higher distress, more likely to be single and unemployed, and less likely to be chronically ill or disabled. Mental health was worse for younger age-groups in each survey year, and this discrepancy is much greater in more recent surveys - consistent with a birth cohort effect. Millennials (those born in the early 1990s) had a lower score at the same age as earlier generations, and the later cohorts do not show the age-related improvement seen in other earlier cohorts as they aged. At age 30 the average MHI-5 score of those born in the 1990s was 67, compared to 72.5 and 74 for people born in the 1980s and 1970s.ConclusionsThe deterioration in mental health over time which has been reported in large cross-sectional surveys, likely reflects cohort-specific effects related to the experiences of young people born in the Millennial generation and, to a lesser extent, those from the immediately prior cohort born in the1980s. We need to understand whether later cohorts are less resilient to similar risk factors experienced by earlier cohorts or whether they experience more and/or a greater severity of risks for mental ill-health. Such evidence is critical if the deteriorating pattern of mental health is to be arrested.Disclosure of InterestNone Declared

An acoustic study on age-related changes in vowel production of Chinese

This paper investigates the relationship between vowel production and age using speech data from 109 Chinese L1 speakers (61 females and 48 males) covering an age range of 20 to 80 years. Acoustical estimation of vocal tract length (VTL) as well as multiple acoustic metrics are analyzed with generalized additive mixed models (GAMM). The results indicate that: (1) After controlling for VTL, F0 and duration, vowels show a centralization trend with increasing age, with a more significant effect observed in female speakers; (2) VTL does not significantly change with age; (3) The patterns observed in vowel distinctiveness and duration may present evidence contradicting the notion of vowel lengthening as a compensatory mechanism; (4) The patterns of age-related changes in different measurements and different genders are diverse. The U-shaped change patterns are found in the male speakers and the age around 50 may serve as a turning point. Based on these findings, we have explored some possible reasons for inconsistent conclusions in previous studies. The physiological aging phenomena of vowel production and potential compensatory mechanisms on motor control abilities, as well as other possible influencing factors, are also discussed.

Early Transport Patterns and Influencing Factors of Different Stocks of Uroteuthis edulis in the East China Sea.

Uroteuthis edulis (U. edulis) is an important economic loliginid resource in the East China Sea (ECS). Its flexible life history traits enable the population to quickly adapt to changes in habitat. Understanding the early transport process helps us to grasp the habitat requirements of populations at key life history stages. In this study, particle tracing was used to simulate the early transport trajectories (within 120 days). The gradient forest method (GFM) and generalized additive mixed models (GAMMs) were used to analyze the key environmental variables that affect the early transport trajectories and the impact of environmental factors on the transport process, respectively. The results showed that spring stock tracers were transported to the northeast of the release area (Pengjiayu water) and the Pacific side of Japan. Summer stock tracers were transported to the north and northeast of the release area (Zhoushan island). Current velocity, salinity, and temperature were key environmental variables that affected the trace element ratios of spring stock at early life history stages. Mixed-layer depth (MLD), velocity, and chlorophyll a concentration (Chla) were key environmental variables for summer stock. Zonal velocity was positively correlated with the trace element ratio for spring and summer stock (0.14-0.16 m/s), while the meridional velocity showed an opposite correlation. The physical driving mechanisms of the Kuroshio warm current (or the Taiwan warm current) and the Yangtze River determine the paralarva retention location during early transportation. The differences in the dominant factors of the water environment in the retention area may affect the paralarva physiological functions and food availability. This study provides a scientific basis for a comprehensive understanding of the migration characteristics of U. edulis with different stocks.

Analysis of the Swordfish Xiphias gladius Linnaeus, 1758 Catches by the Pelagic Longline Fleets in the Eastern Pacific Ocean

During the last 50 years, the increase in the efforts of the longline fleet in the Eastern Pacific Ocean (EPO) resulted in an increase in the capture of the swordfish Xiphias gladius. We analyzed a historical database of swordfish catches (1980–2020) reported by the industrial longline fleet to the Inter-American Tuna Tropical Commission (IATTC), which contains catch and effort data aggregated in monthly quadrants of 5° × 5° in the EPO. The swordfish catch reported by the international longline fleets was analyzed to evaluate the spatiotemporal variation of the catch and the different phases through which this important fishery has gone through. Different statistical models such as the Generalized Additive Mixed Model (GAMM) and the breaks for additive season and trend BFAST algorithm were used for the decomposition of the time series. Results indicated that the effort directed towards the swordfish increased in recent years and that the highest catches occurred by Peru. The adjusted GAMM explained 80% of the total temporal variation of the swordfish catch per unit effort CPUE and had a 90% prediction efficiency. The BFAST algorithm found three break points in the time series of the standardized CPUE, points associated with abrupt changes, thus defining four distinct periods, all of them statistically significant. According to the BFAST model, the current trend of swordfish CPUE is upward. It is recommended to take this finding with caution to obtain the sustainable exploitation of the swordfish fishery resource.

A remote sensing assessment of oak forest recovery after postfire restoration

Mediterranean Europe is experiencing a rise in severe wildfires, resulting in growing socioeconomic and ecological impacts. Postfire restoration has become a crucial approach to mitigate these impacts and promote ecosystem recovery. However, the ecological effects of such interventions are still not well understood. We employed remote sensing techniques to evaluate the impact of postfire emergency stabilization on the recovery of deciduous oak forests in Portugal. Our study encompassed 3013 sampling points located in areas with and without postfire interventions. We chose the Normalized Difference Vegetation Index (NDVI) as an indicator of oak forest recovery over a four-year period following wildfires that took place in 2016 and 2017. We used a Generalized Additive Mixed Model (GAMM) to assess how NDVI changed over time as a function of postfire restoration, fire characteristics, topography, and postfire drought events. We found that postfire restoration had a significant positive effect on NDVI recovery over time, although this effect was small. Severe drought and fire recurrence up to six fires had a negative effect on the recovery of NDVI. Conversely, severe wetness and either low or high burn severities had a positive effect on recovery. Our study emphasizes the importance of monitoring postfire restoration effects on forest recovery to guide restoration planning and improve forest management in burned areas. This becomes even more relevant under increased wildfire severity predicted for the Mediterranean region interacting with other climate-driven disturbances, which will further negatively affect forest recovery.

Effects of gene dosage and development on subcortical nuclei volumes in individuals with 22q11.2 copy number variations.

The 22q11.2 locus contains genes critical for brain development. Reciprocal Copy Number Variations (CNVs) at this locus impact risk for neurodevelopmental and psychiatric disorders. Both 22q11.2 deletions (22qDel) and duplications (22qDup) are associated with autism, but 22qDel uniquely elevates schizophrenia risk. Understanding brain phenotypes associated with these highly penetrant CNVs can provide insights into genetic pathways underlying neuropsychiatric disorders. Human neuroimaging and animal models indicate subcortical brain alterations in 22qDel, yet little is known about developmental differences across specific nuclei between reciprocal 22q11.2 CNV carriers and typically developing (TD) controls. We conducted a longitudinal MRI study in a total of 385 scans from 22qDel (n = 96, scans = 191, 53.1% female), 22qDup (n = 37, scans = 64, 45.9% female), and TD controls (n = 80, scans = 130, 51.2% female), across a wide age range (5.5-49.5 years). Volumes of the thalamus, hippocampus, amygdala, and anatomical subregions were estimated using FreeSurfer, and the linear effects of 22q11.2 gene dosage and non-linear effects of age were characterized with generalized additive mixed models (GAMMs). Positive gene dosage effects (volume increasing with copy number) were observed for total intracranial and whole hippocampus volumes, but not whole thalamus or amygdala volumes. Several amygdala subregions exhibited similar positive effects, with bi-directional effects found across thalamic nuclei. Distinct age-related trajectories were observed across the three groups. Notably, both 22qDel and 22qDup carriers exhibited flattened development of hippocampal CA2/3 subfields relative to TD controls. This study provides novel insights into the impact of 22q11.2 CNVs on subcortical brain structures and their developmental trajectories.

Predicting important fishing grounds for the small-scale fishery, based on Automatic Identification System records, catches, and environmental data

Abstract Effective and sustainable management of small-scale fisheries (SSF) is challenging. We describe a novel approach to identify important fishing grounds for SSF, by implementing a habitat modelling approach, using environmental predictors and Automatic Identification System (AIS)-B data coupled with logbook and First Sales Notes data, within the SE Bay of Biscay. Fishing activity patterns and catches of longliners and netters are used to determine the main environmental characteristics of the fishing grounds, and a habitat modelling approach is implemented to predict the zones that fulfil similar environmental characteristics across a larger geographical extent. Generalized additive mixed models (GAMMs) were built for 24 fish species, and to identify other zones that fulfil similar characteristics and, thus, could be considered relevant for the species targeted by each gear type. Most of the models showed a good prediction capacity. The models included between one and four predictor variables. ‘Depth of mixing layer’ and ‘benthic rocky habitat’ were the variables included more frequently for fish species captured by netter’s fleet. For longliners, the ‘seafloor slope’ and ‘benthic rocky habitat’ were the two most important variables. The predictive maps provide relevant information to assist in management and marine spatial planning.

Spatiotemporal modelling of PM2.5\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\\hbox {PM}_{2.5}$$\\end{document} concentrations in Lombardy (Italy): a comparative study

This study presents a comparative analysis of three predictive models with an increasing degree of flexibility: hidden dynamic geostatistical models (HDGM), generalised additive mixed models (GAMM), and the random forest spatiotemporal kriging models (RFSTK). These models are evaluated for their effectiveness in predicting PM2.5\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ ext {PM}_{2.5}$$\\end{document} concentrations in Lombardy (North Italy) from 2016 to 2020. Despite differing methodologies, all models demonstrate proficient capture of spatiotemporal patterns within air pollution data with similar out-of-sample performance. Furthermore, the study delves into station-specific analyses, revealing variable model performance contingent on localised conditions. Model interpretation, facilitated by parametric coefficient analysis and partial dependence plots, unveils consistent associations between predictor variables and PM2.5\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$\ ext {PM}_{2.5}$$\\end{document} concentrations. Despite nuanced variations in modelling spatiotemporal correlations, all models effectively accounted for the underlying dependence. In summary, this study underscores the efficacy of conventional techniques in modelling correlated spatiotemporal data, concurrently highlighting the complementary potential of Machine Learning and classical statistical approaches.

Freshwater fish otoliths record signals from both water and physiological processes: new insights from Sr/Ca and Ba/Ca ratios

Using strontium (Sr) and barium (Ba) in otoliths to determine natal origins and understand patterns of fish movements is based on the fundamental assumption that otoliths record water chemistry signals without any major alterations. Although prior studies highlighted that fish physiology can modify the water signal in otoliths, studies for freshwater fish are scarce. We exposed different groups of Atlantic salmon parr Salmo salar to different scenarios of ambient-level variations in Sr/Ca and Ba/Ca ratios and then combined otolith chemical profiles with environmental data (water chemistry and temperature), Fulton's index, and otolith growth rates to assess what factors explain/influence the elemental ratios of Sr and Ba in otoliths. Generalized additive mixed models (GAMMs) using water-based otolith composition, temperature, Fulton's index, and “individual” as explanatory variables allow to demonstrate that water chemistry alone cannot fully explain measured ratios in otoliths, except in scenarios involving significant changes in water chemistry. Other factors (physiological effects) should be accounted for reproducing short and minimal seasonal variations in water composition, considering that inter-individual variability contributes quite significantly in most scenarios.

Ontogenetic growth responses to the environmental changes of the swordtip squid (Uroteuthis edulis) spring stock in the East China Sea

AbstractSwordtip squid (Uroteuthis edulis) is characterized by a complex population structure and rapid generation renewal and sensitive to habitat changes. Its population growth response to environmental variations implies its flexible life history traits. In this study, with the samples collected in the north‐central waters of the East China Sea from 2017 to 2021, the daily growth of the spring stock was analyzed based on the age and increment width of statolith. The gradient forest method (GFM) and generalized additive mixed models (GAMMs) were used to explore the changes in the weights of environmental variables and the relationships between daily growth and environmental variables in various growth stages. The age of the samples collected from 2017 to 2021 mainly ranged from 180 to 240 days, and the spring stock was the dominant stock. According to the distance from daily increment to core in the statolith, the life history of the spring stock was divided into four growth stages (S1 embryo–larval stage, S2 juvenile stage, S3 subadult stage, and S4 adult stage). For the spring stock, the cumulative weight of temperature and salinity in the population growth was the largest in S1 stage; the cumulative weight of temperature and velocity was the largest in S2 stage; the cumulative weight of temperature was the largest in S3 stage; and the cumulative weight of mixed layer depth (MLD), temperature, and salinity was the largest in S4 stage. The relationship between temperature at the depth of 25 m (T25) and daily growth of the spring stock was first positively correlated (S1–S2), then negatively correlated (S3), and finally positively correlated (S4). The relationship between environment variable and growth of the spring stock gradually decreased with the increase in MLD (30 to 50 m) and SSS (S3–S4, 32.2‰ to 33.2‰) and gradually increased with the increase in the velocity of currents (S1, .1 to .2 m/s). The differences in the responses of the spring stock to environmental variations in different growth stages may lead to the changes in the growth traits for the spring stock. This study provides a scientific basis for a comprehensive understanding of the life history traits of U. edulis.

Formant dynamics in second language speech: Japanese speakers' production of English liquids.

This article reports an acoustic study analysing the time-varying spectral properties of word-initial English liquids produced by 31 first-language (L1) Japanese and 14 L1 English speakers. While it is widely accepted that L1 Japanese speakers have difficulty in producing English /l/ and /ɹ/, the temporal characteristics of L2 English liquids are not well-understood, even in light of previous findings that English liquids show dynamic properties. In this study, the distance between the first and second formants (F2-F1) and the third formant (F3) are analysed dynamically over liquid-vowel intervals in three vowel contexts using generalised additive mixed models (GAMMs). The results demonstrate that L1 Japanese speakers produce word-initial English liquids with stronger vocalic coarticulation than L1 English speakers. L1 Japanese speakers may have difficulty in dissociating F2-F1 between the liquid and the vowel to a varying degree, depending on the vowel context, which could be related to perceptual factors. This article shows that dynamic information uncovers specific challenges that L1 Japanese speakers have in producing L2 English liquids accurately.

Spatial distribution of microplastics in a coastal upwelling region: Offshore dispersal from urban sources in the Humboldt Current System

In coastal waters, higher concentrations of microplastics (MPs) are generally related to densely populated and industrialized areas, but intense upwelling and offshore transport in the Eastern Boundary Upwelling Systems (EBUS) may influence this pattern. The Humboldt Current System (HCS) along the coast of northern-central Chile represents a perfect model to test whether the abundance of MP at the sea surface decreases with distance from land-based sources, e.g., river mouths, harbors, and submarine wastewater outfalls. The sea surface was sampled with a manta trawl to examine the abundance, composition, and distribution of floating MPs, and Generalized Additive Mixed Models (GAMMs) were performed to examine the relationship between MP abundance (particles km−2) and the distance to putative sources. MPs were found in all 57 net tows, with an average of ⁓120,000 MP km−2 and maximum values of ⁓1,500,000 MP km−2. The composition of MPs was dominated by fragments (>50% of the total count) and over 80% of all MPs were ≥1 mm. The combined effect of the various sources, spatially concentrated in urban areas, makes it difficult to distinguish their relative contributions, but the MP composition suggested that rivers are more important sources, followed by submarine wastewater outfalls and then harbors. A significant and steep negative relationship with the “distance to source” explained 15.2% of the variance of “MP abundance”, suggesting rapid offshore displacement within the HCS. This is the first study to report this pattern along the edges of the South Pacific Subtropical Gyre (SPSG), revealing that continuous offshore transport of microplastic from land-based sources is occurring over large scales and contributing to the accumulation of microplastics in the center of the SPSG. However, the findings additionally suggested that processes at meso- and submeso-spatial scales (driven by geographic and seasonal variables) are disrupting the general pattern.

Low ambient temperature exposure increases the risk of ischemic stroke by promoting platelet activation

BackgroundAccumulating epidemiological evidence suggests the association between low ambient temperature exposure and the risk of ischemic stroke, but the underlying mechanisms remain unclear. ObjectiveGiven the crucial role of platelet activation and thrombosis in ischemic stroke, this study aims to investigate the effect of ambient temperature on platelet activation through multi-center clinical data in Tianjin as well as animal experiments. MethodsFrom 2018 to 2020, nearly 3000 ischemic stroke patients from three stroke centers in Tianjin were included in the analysis, among them the ADP induced platelet aggregation rate was available. Meteorological data from the same period had also been collected. After controlling for confounding factors, the generalized additive mixed model (GAMM) was used to evaluate the correlation between environmental temperature and platelet aggregation rate. In further animal experiments, platelet function assessments were conducted on mice from the cold exposure group and the normal temperature group, including platelet aggregation, spreading, and clot retraction. Additionally, tail bleeding and mesentery thrombosis were also tested to monitor hemostasis and thrombosis in vivo. ResultA nonlinear “S” shaped relationship between outdoor temperature and platelet aggregation was found. Each 1 °C decrease of mean temperature was associated with an increase of 7.77 % (95 % CI: 2.06 % - 13.48 %) in platelet aggregation. The ambient temperature is not related to other platelet parameters. Subgroup analysis found that males, people aged ≥65 years, and hypertensive individuals are more susceptible to temperature changes. Furthermore, animal experiments demonstrated that the increased CIRBP levels and subsequent activation of p-AKT/p-ERK may be one of the reasons for cold exposure induced platelets activation. ConclusionBoth clinical data and basic research support that low ambient temperature exposure has the potential to increase platelet activation. These results provide a basis for understanding the potential mechanism of temperature variations on the pathogenesis of cerebrovascular diseases.

Sexual dimorphisms in longitudinal associations of the blood lipidome with cognitive decline in Alzheimer’s disease

AbstractBackgroundLate‐onset Alzheimer’s disease (AD) is the leading cause of dementia with neither cure nor a clearly understood disease trajectory. Previous studies have shown cross‐sectional and longitudinal associations of the blood lipidome and AD. We previously reported on significant sex differences in the associations of lipids with AD biomarkers and it has been suggested that such sex differences may result in distinct trajectories of AD progression. We performed an association study of sex‐stratified longitudinal lipidomics data with cognitive function to further characterize sexual dimorphisms in lipidomic manifestations of AD.MethodWe used a subset of the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset: 1349 individuals (745 males and 604 females) with blood samples and clinical data collected at 10 time points with intervals of 6 to12 months (4145 samples total). We analyzed 749 lipid species measured by quantitative LC‐MS lipidomics profiling (Baker Heart and Diabetes Institute, Melbourne, Australia). We calculated linear mixed models (LMMs) and generalized additive mixed models (GAMMs) to analyze time‐resolved associations between lipids and two composite cognitive scores (ADNI memory and executive function) while adjusting for age, sex (pooled analysis only), BMI, HDL‐C, total cholesterol, copies of APOE ε4, years of education as well as significant medication effects.ResultWe observed a significant difference in lipid concentrations between males and females in line with previous studies, however, these differences were not fully consistent across diagnostic groups. In longitudinal association analyses, we found a substantial fraction of lipid species to have significant sex‐specific associations with cognitive function, which were not observed in the pooled analysis. These lipids were distributed across several lipid classes with enrichment in phosphatidylcholines (PC), phosphatidylethanolamines (PE) and triglycerides (TG). Comparisons between LMMs and GAMMs further indicated that some lipids might have a nonlinear relationship with cognitive trajectories.ConclusionOur results suggest that there are substantial sex differences in longitudinal associations of the blood lipidome with cognitive decline resulting in complex and partially nonlinear relationships. The highlighted lipid pathways can serve as a starting point to better characterize sex‐specific molecular changes linked to cognitive decline in AD.

Prognostic Value of the Baseline and Early Changes in Monocyte-to-Lymphocyte Ratio for Short-Term Mortality among Critically Ill Patients with Acute Kidney Injury

(1) Background: Inflammation plays an important role in the onset and progression of acute kidney injury (AKI). Despite this, evidence regarding the prognostic effect of the monocyte-to-lymphocyte ratio (MLR), a novel systemic inflammation marker, among patients with AKI is scarce. This study sets out to investigate the prognostic potential of both baseline and early changes in MLR for short-term mortality among critically ill patients with AKI. (2) Method: Eligible patients with AKI from the Medical Information Mart for Intensive Care IV database were retrospectively analyzed. MLR cutoff values were determined using maximally selected rank statistics and tertiles. The clinical outcomes were 30-day and 90-day mortality in the intensive care unit. A restricted cubic splines model and Cox proportional hazards models were utilized to evaluate the association between the baseline MLR and short-term mortality. Then, the trends in MLR over time were compared between the 30-day survivors and non-survivors using a generalized additive mixed model (GAMM). (3) Result: A total of 15,986 patients were enrolled. Multivariable Cox regression analysis identified baseline MLR ≥ 0.48 as an independent risk factor predicting 30-day mortality (HR 1.33, 95%CI 1.24, 1.45, p < 0.001) and 90-day mortality (HR 1.34, 95%CI 1.23, 1.52, p < 0.001) after adjusting for potential confounders. Similar trends were observed for 30-day and 90-day mortality when tertiles were used to group patients. The restricted cubic splines model revealed a non-linear association between MLR and 30-day and 90-day mortality (both p for non-linear < 0.001, both p for overall < 0.001). The area under the curve of 0.64 for MLR was higher than that of monocytes (0.55) and lymphocytes (0.61). In the subgroup analyses, despite the noted significant interactions, the direction of the observed association between MLR and 30-day mortality was consistent across most prespecified subgroups, except for shock and black ethnicity. The GAMM results highlighted that, as time went on, MLR in the 30-day survival group consistently declined, whereas MLR in the non-survival group rose within 15 days post-ICU admission. The difference between the two groups persisted significantly even after adjusting for confounders (p = 0.006). (4) Conclusion: A higher baseline MLR was identified as an independent risk factor predicting 30-day and 90-day mortality. The early increase in MLR was associated with high 30-day mortality, suggesting that dynamic monitoring of MLR could potentially better predict survival in critically ill patients with AKI.

The overlooked effect of amplitude on within-speaker vowel variation

Abstract We analyse variation in vowel production within monologues produced by speakers in a quiet, well-controlled environment. Using principal component analysis (PCA) and generalized additive mixed models (GAMMs), applied to a large corpus of naturalistic recordings of New Zealand English speakers, we show that the first formant of monophthongs varies significantly with variation in a speaker’s relative amplitude. We also find that amplitude variation is used, potentially agentively, to mark the beginning and ending of topical sections within single-speaker monologues. These results have significant methodological consequences for the study of vocalic variation in the context of research on speaker style and language variation and change. While laboratory research has shown a connection between variation in F1 and amplitude in loud environments or with distant interlocutors, this has not been seen in quiet environments with unscripted speech of the sort often used in sociolinguistcs. We argue that taking account of this variation is an important challenge for both within-speaker investigation of stylistic covariation and across-speaker investigation. In the latter case we recommend, as a minimal step, the inclusion of a measure of relative amplitude within regression models.

Optimizing Refractive Outcomes of SMILE: Artificial Intelligence versus Conventional State-of-the-Art Nomograms

Purpose AI (artificial intelligence)-based methodologies have become established tools for researchers and physicians in the entire field of ophthalmology. However, the potential of AI to optimize the refractive outcome of keratorefractive surgery by means of machine learning (ML)-based nomograms has not been exhausted yet. In this study, we wanted to comprehensively compare state-of-the-art conventional nomograms for Small-Incision-Lenticule-Extraction (SMILE) with a novel ML-based nomogram regarding both their spherical and astigmatic predictability. Methods A total of 1,342 eyes were analyzed for creation of three different nomograms based on a linear model (LM), a generalized additive mixed model (GAMM) and an artificial-neuronal-network (ANN), respectively. A total of 16 patient- and treatment-related features were included. Each model was trained by 895 eyes and validated by the remaining 447 eyes. Predictability was assessed by the difference between attempted and achieved change in spherical equivalent (SE) and the difference between target induced astigmatism (TIA) and surgically induced astigmatism (SIA). The root mean squared error (RMSE) of each model was computed as a measure of overall model performance. Results The RMSE of LM, GAMM and ANN were 0.355, 0.348 and 0.367 for the prediction of SE and 0.279, 0.278 and 0.290 for the astigmatic correction, respectively. By applying the created models, the theoretical yield of eyes within ±0.50 D of SE from target refraction improved from 82 to 83% (LM), 84% (GAMM) and 83% (ANN), respectively. Astigmatic outcomes showed an improvement of eyes within ±0.50 D from TIA from 90 to 93% (LM), 93% (GAMM) and 92% (ANN), respectively. Subjective manifest refraction was the single most influential covariate in all models. Conclusion Machine learning endorsed the validity of state-of-the-art linear and non-linear SMILE nomograms. However, improving the accuracy of subjective manifest refraction seems warranted for optimizing ±0.50 D SE predictability beyond an apparent methodological 90% limit.

Higher live birth rates are associated with a normal body mass index in preimplantation genetic testing for aneuploidy frozen embryo transfer cycles: a Society for Assisted Reproductive Technology Clinic Outcome Reporting System study

To determine whether body mass index (BMI) was associated with live birth in patients undergoing transfer of frozen-thawed preimplantation genetic testing for aneuploidy (PGT-A) embryos. Retrospective cohort study of cycles reported to the Society for Assisted Reproductive Technology Clinic Outcome Reporting System. All autologous and donor recipient PGT-A-tested cycles reported to the Society for Assisted Reproductive Technology Clinic Outcome Reporting System from 2014 to2017. Body mass index. The primary outcome measure was the live birth rate, and the secondary outcome measures were the clinical pregnancy and biochemical pregnancy rates. Multivariable generalized additive mixed models and log-binomial models were used to model the relationship between BMI and outcome measures. A total of 77,018 PGT-A cycles from 55,888 patients were analyzed. Of these cycles, 70,752 were autologous, and 6,266 were donor recipient. In autologous cycles, a statistically significant and clear nonlinear relationship was observed between the BMI and live birth rates, with the highest birth rates observed for the BMI range of 23-24.99 kg/m2. When using 23-24.99 kg/m2 as the referent, other BMI ranges demonstrated a lower probability of live birth and clinical pregnancy that continued to decrease as the BMI moved further from the reference value. Patients with a BMI of <18.5 kg/m2 had a 11% lower probability of live birth, whereas those with a BMI of ≥40 kg/m2 had a 27% lower probability than the referent. A normal-weight BMI range of 23-24.99 kg/m2 was associated with the highest probability of clinical pregnancy and live birth after a frozen-thawed PGT-A-tested blastocyst transfer in both autologous and donor recipient cycles. A BMI outside the range of 23-24.99 kg/m2 is likely associated with a malfunction in the implantation process, which is presumed to be related to a uterine factor and not an oocyte factor, as both autologous and donor recipient cycle outcomes were associated similarly with the BMI of the intended parent.

Privacy-preserving and lossless distributed estimation of high-dimensional generalized additive mixed models

Various privacy-preserving frameworks that respect the individual’s privacy in the analysis of data have been developed in recent years. However, available model classes such as simple statistics or generalized linear models lack the flexibility required for a good approximation of the underlying data-generating process in practice. In this paper, we propose an algorithm for a distributed, privacy-preserving, and lossless estimation of generalized additive mixed models (GAMM) using component-wise gradient boosting (CWB). Making use of CWB allows us to reframe the GAMM estimation as a distributed fitting of base learners using the L2\\documentclass[12pt]{minimal} \\usepackage{amsmath} \\usepackage{wasysym} \\usepackage{amsfonts} \\usepackage{amssymb} \\usepackage{amsbsy} \\usepackage{mathrsfs} \\usepackage{upgreek} \\setlength{\\oddsidemargin}{-69pt} \\begin{document}$$L_2$$\\end{document}-loss. In order to account for the heterogeneity of different data location sites, we propose a distributed version of a row-wise tensor product that allows the computation of site-specific (smooth) effects. Our adaption of CWB preserves all the important properties of the original algorithm, such as an unbiased feature selection and the feasibility to fit models in high-dimensional feature spaces, and yields equivalent model estimates as CWB on pooled data. Next to a derivation of the equivalence of both algorithms, we also showcase the efficacy of our algorithm on a distributed heart disease data set and compare it with state-of-the-art methods.