Latent Model Research Articles

Patient-reported physical well-being predicts good long-term survival of hematopoietic stem cell transplantation.

This study aimed to explore the association between patient-reported items at different time points after hematopoietic stem cell transplantation (HSCT) and long-term survival. We conducted a study with 144 allogeneic HSCT patients, following them for 5 years post-transplantation. Data from the Functional Assessment of Cancer Therapy-Bone Marrow Transplant (FACT-BMT) questionnaire were collected before transplantation and at 1, 3, 6, 12, 18, 36, and 60 months after transplantation. Demographic characteristics and survival status were also assessed. Among the 144 cases, the 5-year overall survival (OS), progression-free survival (PFS), non-relapse mortality (NRM), and graft-versus-host disease-free (GRFS) rates were 65%, 48%, 17%, and 36% respectively. Health-related quality of life (HRQOL) showed a fluctuating pattern over 5 years. Using a latent class mixed model, patients were classified into two groups based on their physical well-being (PWB) scores during the 60-month follow-up. Class 1 had initially lower PWB scores, which gradually increased over time. In contrast, Class 2 maintained higher PWB scores with slight increases over time. Kaplan-Meier survival analysis revealed that Class 1 had better OS (70.9% vs. 52.9%, p = 0.021), PFS (60.5% vs. 41.2%, p = 0.039), and GRFS (35.1% vs. 29.3%, p = 0.035) compared to Class 2. Patients who had higher initial PWB scores after HSCT demonstrated improved long-term survival outcomes. The PWB score could serve as a valuable predictor for the prognosis of HSCT.

A data-driven analysis to determine the optimal number of topics 'K' for latent Dirichlet allocation model

Topic modeling is an unsupervised machine learning technique successfully used to classify and retrieve textual data. However, the performance of topic models is sensitive to selecting optimal hyperparameters, the number of topics 'K' and Dirichlet priors 'α' and 'β.' This data-driven analysis aims to determine the optimum number of topics, 'K,' within the latent Dirichlet allocation (LDA) model. This work utilizes three datasets, namely 20-Newsgroups news articles, Wikipedia articles, and Web of Science containing science articles, to assess and compare various 'K' values through the grid search approach. The grid search approach finds the best combination of hyperparameter values by trying all possible combinations to see which performs best. This research seeks to identify the 'K' that optimizes topic relevance, coherence, and model performance by leveraging statistical metrics, such as coherence scores, perplexity, and topic distribution quality. Through empirical analysis and rigorous evaluation, this work provides valuable insights for determining the ideal 'K' for LDA models.

Barriers to Participation in Organized Physical Activity Among LGBTQ+ Youth: Differences by Sexual, Gender, and Racial Identities.

LGBTQ+ youth engage in organized physical activity to a lesser degree than their cisgender and heterosexual counterparts. Existing literature on this organized physical activity disparity is limited, particularly with LGBTQ+ youth samples. The current analysis examined individual and systemic barriers to organized physical activity for LGBTQ+ youth across sexual, gender, and racial identities. A subsample of LGBTQ+ students (N = 4566) from the 2021 Dane County Youth Assessment completed items that measured barriers to organized physical activity and systemic factors (ie,family money problems and bias-based bullying) associated with access to organized physical activity. Latent class analysis discerned patterns of individual and systemic barriers to organized physical activity. Latent class regression modeling tested gender, sexual, and racial identities as correlates of latent class membership. More than half of the sample did not participate in organized physical activity. Four profiles of LGBTQ+ youth were discerned based on self-reported barriers: high barrier (8%), bullied (16%), low interest or perceived skills (28%), and low barrier (48%). The low-barrier class included a greater proportion of LGBTQ+ youth who identified as White, or cisgender, or heterosexual as well as youth self-reporting higher organized physical activity. The high-barrier and bullied classes comprised more marginalized gender and sexual identities. LGBTQ+ youth experience individual and systemic barriers to organized physical activity, including inequitable access and bullying, and barriers are uniquely experienced across sexual, gender, and racial identities. Physical activity promotion among LGBTQ+ youth would be strengthened by policies that address inequitable access to opportunities and bias-based bullying.

Profiling people with Parkinson's disease at risk of cognitive decline: Insights from PPMI and ICICLE-PD data.

A subset of people with Parkinson's disease (PD) develop dementia faster than others. We aimed to profile PD cognitive subtypes at risk of dementia based on their rate of cognitive decline. Latent class mixed models stratified subtypes in Parkinson's Progression Markers Initiative (PPMI) (N=770) and ICICLE-PD (N=212) datasets based on their decline in the Montreal Cognitive Assessment over at least 4 years. Baseline demographic and cognitive data at diagnosis were compared between subtypes to determine their clinical profile. Four subtypes were identified: two with stable cognition, one with steady decline, and one with rapid decline. Performance on Judgement of Line Orientation, but not category fluency, was associated with a steady decline in the PPMI dataset, and deficits in category fluency, but not visuospatial function, were associated with a steady decline in the ICICLE-PD dataset. People with PD susceptible to cognitive decline demonstrate unique clinical profiles at diagnosis, although this differed between cohorts. Four cognitive subtypes were revealed in two Parkinson's disease samples.Unique profiles of cognitive impairment were related to cognitive decline.Judgement of Line Orientation/category fluency predictive of steady decline.Global deficits related to rapid cognitive decline and increased dementia risk.

Distinguishing sympathisers, philanthropists, rusted on activists and radicals: Using person‐centred analyses in collective action research

AbstractMany of our theories and models within social psychology implicitly refer to ‘types’ or subgroups of people who engage in collective action (activists, protestors), or do not (sympathisers, bystanders). Other frameworks differentiate between actors based on their adoption of tactics (benevolents, activists, radicals). How, empirically, do we distinguish sympathisers from activists? Activists from radicals? This paper describes recent research that adopts person‐centred statistical approaches (e.g., latent profile analysis, latent transition analysis, latent growth mixture models) to address three contemporary puzzles of collective action research. We present research showing that the methods are useful in identifying and explaining sub‐groups of people who seek to bring about change in qualitatively different ways. The methods are also useful for understanding volatility in collective action as well as identifying and understanding variation in how people sustain, increase, or diminish their commitment over time. There is nothing so practical as a good method and person‐centred statistical approaches are an important complement to variable‐centred approaches in social psychological research on collective action.

Structural characteristics and non-linear fiscal multipliers

We identify structural characteristics of economies which determine the fiscal multipliers in normal times. We find that structural variables determine fiscal multipliers in a non-linear fashion where fiscal multipliers change significantly across the threshold of structural variables. Using the recently developed Panel Latent Threshold Model, we identify latent groups of countries intrinsic to the structural characteristics of the variables. Results suggest that improvement in human development and financial infrastructure measures play a prominent role in enhancing the fiscal multiplier effects across the economies. We also find that economies can achieve higher fiscal multipliers by keeping a check on the import propensity. We identify female labour force participation as one of the significant structural determinants of the fiscal multiplier, where many latent groups find parallels with the declining portion of the feminisation U-shaped hypothesis. Overall, the results suggest that countries should pursue targeted structural change-oriented fiscal policy to achieve higher economic growth.

Long-Term Depressive Symptom Trajectories and Midlife Cognition: The CARDIA Study.

The nature of associations between depressive symptoms and cognition early in the life course remains unclear, and racial differences in these associations are not well characterized. The aim of this study was to examine the relationship between trajectories of depressive symptom over 20 years, beginning in young adulthood, and cognitive functions in middle-age among Black and White adults. We used prospective data from participants of the Coronary Artery Risk Development in Young Adults Study. Depressive symptoms were measured at 5 study visits between 1990 and 2010 using the Center for Epidemiologic Studies Depression scale. We used latent class group-based modeling to identify 4 trajectories: "persistently low," "persistently medium," "medium decreasing," and "high increasing" depressive symptoms. In 2015, cognitive function was measured using the Digit Symbol Substitution Test (DSST), Stroop test (reverse coded), and Rey Auditory-Verbal Learning Test (RAVLT).We excluded participants who missed the cognitive battery or had no depressive symptoms measurements, resulting in a total of 3,117 participants. All cognitive tests were standardized, and linear regression was used to relate depressive trajectories with 2015 cognitive functions. The mean [SD] baseline age was 30.1 [3.6] years, and 57% were female. The associations between depressive symptoms and cognition significantly differed by race (p < 0.05). Among Black individuals, compared with having "persistently low," having "medium decreasing," "persistently medium," or "high increasing" depressive symptoms were associated with worse verbal memory, processing speed, and executive function scores (RAVLT persistently medium vs low: β = -0.30, 95% CI -0.48 to -0.12; and high increasing vs low: β = -0.49, 95% CI -0.70 to -0.27; DSST persistently medium vs low: β = -0.28, 95% CI -0.47 to -0.09; and high increasing vs low: β = -0.64, 95% CI -0.87 to -0.42; Stroop persistently medium vs low: β = -0.46, 95% CI -0.70 to -0.23; and high increasing vs low: β = -0.76, 95% CI -1.04 to -0.47). Associations were slightly weaker among White individuals, but we still found that having 'high increasing' depressive symptoms was associated with worse verbal memory and processing speed scores (high increasing vs low: β = -0.38, 95% CI -0.61 to -0.15; and β = -0.40, 95% CI -0.63 to -0.18, respectively). Prolonged exposure to elevated depressive symptoms beginning in young adulthood may result in worse cognitive function over midlife. This association was stronger among Black adults.

Preschool Morphological Awareness and Developmental Change in Early Reading Ability

ABSTRACT Purpose This longitudinal study examined the contribution of preschool morphological awareness to word reading skills and reading comprehension, as well as to the developmental change of reading ability beyond other well-established oral language and cognitive predictors. A distinction was made between the domains of inflectional and derivational morphology. Method Two hundred and fifty-nine Norwegian-speaking children (46% female) with a mean age of 5.5 years were assessed in preschool on language measures and again in Grades 1 and 3 on measures of word reading accuracy and fluency and in Grades 3 and 4 on reading comprehension. We fit latent change score models with preschool predictors using parceling to control for measurement error. Results We found a unique contribution of preschool morphological awareness to reading comprehension in Grade 3, but no unique contribution to Grade 1 decoding. Neither awareness of inflections nor awareness of derivations predicted developmental change in word reading fluency between Grades 1 and 3 or change in reading comprehension between Grades 3 and 4 beyond the effect of control variables. Conclusion Our findings confirm the relevance of morphological awareness only for early attainment in reading comprehension and highlight the importance of accounting for measurement error in studying associations among variables aiming to discover specific contributions.

"Do not inject our babies": a social listening analysis of public opinion about authorizing pediatric COVID-19 vaccines.

Designing effective childhood vaccination counseling guidelines, public health campaigns, and school-entry mandates requires a nuanced understanding of the information ecology in which parents make vaccination decisions. However, evidence is lacking on how best to "catch the signal" about the public's attitudes, beliefs, and misperceptions. In this study, we characterize public sentiment and discourse about vaccinating children against SARS-CoV-2 with mRNA vaccines to identify prevalent concerns about the vaccine and to understand anti-vaccine rhetorical strategies. We applied computational topic modeling to 149 897 comments submitted to regulations.gov in October 2021 and February 2022 regarding the Food and Drug Administration's Vaccines and Related Biological Products Advisory Committee's emergency use authorization of the COVID-19 vaccines for children. We used a latent Dirichlet allocation topic modeling algorithm to generate topics and then used iterative thematic and discursive analysis to identify relevant domains, themes, and rhetorical strategies. Three domains emerged: (1) specific concerns about the COVID-19 vaccines; (2) foundational beliefs shaping vaccine attitudes; and (3) rhetorical strategies deployed in anti-vaccine arguments. Computational social listening approaches can contribute to misinformation surveillance and evidence-based guidelines for vaccine counseling and public health promotion campaigns.

Retrieving and reconstructing conceptually similar images from fMRI with latent diffusion models and a neuro-inspired brain decoding model

Objective. Brain decoding is a field of computational neuroscience that aims to infer mental states or internal representations of perceptual inputs from measurable brain activity. This study proposes a novel approach to brain decoding that relies on semantic and contextual similarity. Approach. We use several functional magnetic resonance imaging (fMRI) datasets of natural images as stimuli and create a deep learning decoding pipeline inspired by the bottom-up and top-down processes in human vision. Our pipeline includes a linear brain-to-feature model that maps fMRI activity to semantic visual stimuli features. We assume that the brain projects visual information onto a space that is homeomorphic to the latent space of last layer of a pretrained neural network, which summarizes and highlights similarities and differences between concepts. These features are categorized in the latent space using a nearest-neighbor strategy, and the results are used to retrieve images or condition a generative latent diffusion model to create novel images. Main results. We demonstrate semantic classification and image retrieval on three different fMRI datasets: Generic Object Decoding (vision perception and imagination), BOLD5000, and NSD. In all cases, a simple mapping between fMRI and a deep semantic representation of the visual stimulus resulted in meaningful classification and retrieved or generated images. We assessed quality using quantitative metrics and a human evaluation experiment that reproduces the multiplicity of conscious and unconscious criteria that humans use to evaluate image similarity. Our method achieved correct evaluation in over 80% of the test set. Significance. Our study proposes a novel approach to brain decoding that relies on semantic and contextual similarity. The results demonstrate that measurable neural correlates can be linearly mapped onto the latent space of a neural network to synthesize images that match the original content. These findings have implications for both cognitive neuroscience and artificial intelligence.

Associations between sleep duration trajectories and risk of cardio-metabolic disease among middle-aged and older Chinese adults

BackgroundThe association of a single time-point measure of sleep duration with cardio-metabolic disease has been extensively studied, but few studies have focused on the impact of sleep duration trajectory. This study aims to model the sleep duration trajectory as predictors for the subsequent development of cardio-metabolic disease. MethodsThis study recruited a notably large population (n = 9883) of subjects aged at least 45 years from the China Health and Retirement Longitudinal Study (CHARLS), who participated in sequential surveys conducted in 2011, 2013, 2015, and 2018. Sleep duration trajectories were plotted using data of night sleep duration recorded at intervals from 2011 to 2015 by latent class trajectory model. The onset of cardio-metabolic diseases from 2015 to 2018 were confirmed and then the risk of different sleep duration trajectories on incident cardio-metabolic disease was examined using cox proportional hazards regression model. ResultsWe identified four sleep duration trajectories. Compared to the normal-stable trajectory, the short-stable trajectory was significantly associated with higher risk of incident stroke (hazard ratio [HR], 1.32; 95 % confidence interval [CI], 1.02 to 1.70), dyslipidemia (HR, 1.22; 95%CI, 1.01 to 1.49), and diabetes (HR, 1.42; 95%CI, 1.13 to 1.78) within three years of follow-up, and the short-increasing trajectory predicted a higher risk of incident stroke (HR, 2.38; 95%CI, 1.25 to 4.55). ConclusionsShort sleep trajectory could increase the risk of incident stroke, dyslipidemia, and diabetes, and an increasing sleep trajectory was associated with increased risk of incident stroke among middle-aged and older Chinese adults.

The genetic architecture of biological age in nine human organ systems.

Investigating the genetic underpinnings of human aging is essential for unraveling the etiology of and developing actionable therapies for chronic diseases. Here, we characterize the genetic architecture of the biological age gap (BAG; the difference between machine learning-predicted age and chronological age) across nine human organ systems in 377,028 participants of European ancestry from the UK Biobank. The BAGs were computed using cross-validated support vector machines, incorporating imaging, physical traits and physiological measures. We identify 393 genomic loci-BAG pairs (P < 5 × 10-8) linked to the brain, eye, cardiovascular, hepatic, immune, metabolic, musculoskeletal, pulmonary and renal systems. Genetic variants associated with the nine BAGs are predominantly specific to the respective organ system (organ specificity) while exerting pleiotropic links with other organ systems (interorgan cross-talk). We find that genetic correlation between the nine BAGs mirrors their phenotypic correlation. Further, a multiorgan causal network established from two-sample Mendelian randomization and latent causal variance models revealed potential causality between chronic diseases (for example, Alzheimer's disease and diabetes), modifiable lifestyle factors (for example, sleep duration and body weight) and multiple BAGs. Our results illustrate the potential for improving human organ health via a multiorgan network, including lifestyle interventions and drug repurposing strategies.

Heterogeneity of treatment effect of vilobelimab in COVID-19: a secondary analysis of a randomised controlled trial

In a phase 3 trial (PANAMO, NCT04333420), vilobelimab, a complement 5a (C5a) inhibitor, reduced 28-day mortality in mechanically ventilated COVID-19 patients. This post hoc analysis of 368 patients aimed to explore treatment heterogeneity through unsupervised learning. All available clinical variables at baseline were used as input. Treatment heterogeneity was assessed using latent class analysis (LCA), Ward’s hierarchical clustering (HC) and the adjudication to previously described clinical sepsis phenotypes. The primary outcome was 28-day mortality. For LCA, a 2-class latent model was deemed most suitable. In the LCA model, 82 (22%) patients were assigned to class 1 and 286 (78%) to class 2. Class 1 was defined by more severely ill patients with significantly higher mortality. In an adjusted logistic regression, no heterogeneity of treatment effect (HTE) between classes was observed (p = 0.998). For HC, no significant classes were found (p = 0.669). Using the previously described clinical sepsis subtypes, 41 patients (11%) were adjudicated subtype alpha (α), 17 (5%) beta (β), 112 (30%) delta (δ) and 198 (54%) gamma (γ). HTE was observed between clinical subtypes (p = 0.001) with improved 28-day mortality after treatment with vilobelimab for the δ subtype (OR = 0.17, 95% CI 0.07–0.40, p < 0.001). No signal for harm of treatment with vilobelimab was observed in any class or clinical subtype. Overall, treatment effect with vilobelimab was consistent across different classes and subtypes, except for the δ subtype, suggesting potential additional benefit for the most severely ill patients.

A geospatial analysis of local intermediate snail host distributions provides insight into schistosomiasis risk within under-sampled areas of southern Lake Malawi

BackgroundAlong the southern shoreline of Lake Malawi, the incidence of schistosomiasis is increasing with snails of the genera Bulinus and Biomphalaria transmitting urogenital and intestinal schistosomiasis, respectively. Since the underlying distribution of snails is partially known, often being focal, developing pragmatic spatial models that interpolate snail information across under-sampled regions is required to understand and assess current and future risk of schistosomiasis.MethodsA secondary geospatial analysis of recently collected malacological and environmental survey data was undertaken. Using a Bayesian Poisson latent Gaussian process model, abundance data were fitted for Bulinus and Biomphalaria. Interpolating the abundance of snails along the shoreline (given their relative distance along the shoreline) was achieved by smoothing, using extracted environmental rainfall, land surface temperature (LST), evapotranspiration, normalised difference vegetation index (NDVI) and soil type covariate data for all predicted locations. Our adopted model used a combination of two-dimensional (2D) and one dimensional (1D) mapping.ResultsA significant association between normalised difference vegetation index (NDVI) and abundance of Bulinus spp. was detected (log risk ratio − 0.83, 95% CrI − 1.57, − 0.09). A qualitatively similar association was found between NDVI and Biomphalaria sp. but was not statistically significant (log risk ratio − 1.42, 95% CrI − 3.09, 0.10). Analyses of all other environmental data were considered non-significant.ConclusionsThe spatial range in which interpolation of snail distributions is possible appears < 10km owing to fine-scale biotic and abiotic heterogeneities. The forthcoming challenge is to refine geospatial sampling frameworks with future opportunities to map schistosomiasis within actual or predicted snail distributions. In so doing, this would better reveal local environmental transmission possibilities.Graphical

Autonomy-supportive parenting as a moderator of associations between ethnic and racial identity processes and psychological adjustment during college.

The purpose of this study was to test the roles of ethnic and racial identity (ERI) processes and autonomy-supportive parenting on college students' psychological adjustment. American college students of color (N = 505) completed questionnaires assessing ERI exploration and commitment, autonomy-supportive parenting, and psychological adjustment (self-esteem, depressive symptoms). Key variables were operationalized as latent constructs, and main and interaction effects were tested using the latent moderated structural equation modeling approach. Higher levels of ERI commitment (but not exploration) and parental autonomy support each uniquely predicted higher levels of self-esteem and lower levels of depressive symptoms. Parental autonomy support moderated associations between ERI processes and psychological adjustment, and the nature of moderation did not differ across Black and Latino/a/x students. Supporting the psychological adjustment of college students of color necessitates acknowledging the importance of both parental and institutional efforts to encourage students' autonomy strivings and ERI processes. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Identifying subtypes in persons, situations and person-situation interactions: Categorical latent state-trait modelling approaches.

The latent state-trait theory posits that a psychological construct may reflect stable influences specific to a person (i.e., trait), ephemeral influences from situations (i.e., state), and interactions between them (i.e., state-trait interactions). Researchers conventionally apply mixture modelling to explore heterogeneity in variables by identifying homogenous classes with respect to the measured variable, yet rarely distinguishing between person- and situation-specific classes. The current study introduces novel categorical latent state-trait models to identify subgroups in states and traits, quantifying the effects of person-specific classes, situation-specific classes, and person-situation interactions. The proposed models are applied to an empirical dataset. We discuss statistical inference, effect size measures, and model visualization for the proposed models. Based on realistic parameter values from the empirical dataset, preliminary simulation studies were conducted to investigate models' performances. Bayesian estimation in the proposed models allows flexible testing of a wide range of hypotheses related to state, trait, and interaction effects. We discuss limitations and future directions.

Developmental pathways from insomnia to suicidality: A resilience perspective

BackgroundInsomnia has been identified as a potential risk factor for suicidality. However, to date, few studies using the longitudinal-design have examined the underlying mechanism of this relationship. Based on a resilience perspective, this study aimed to investigate the dynamic longitudinal relationships between insomnia, resilience, and suicidality. MethodsA total of 5785 freshmen were sampled from a large-scale health-related cohort among Chinese college students. This study spanned six waves, covering the period from 2020 to 2022. Data from T1 to T4 were used because resilience was not measured at baseline (T0) and T5. The cross-lagged panel models and the latent growth curve mediation model were used to examine the longitudinal dynamic relationships between insomnia, resilience, and suicidality. ResultsThe results showed that insomnia symptoms and suicidality mutually predicted each other, and resilience played a longitudinal mediating role in linking insomnia symptoms and suicidality. ConclusionsGiven that resilience served as a mediator in the relation between insomnia symptoms and suicidality, some resilience-oriented prevention and intervention programs will be helpful in reducing the risk of suicide among university students.

A study of latent profile analysis of empathic competence and factors influencing it in nursing interns: a multicenter cross-sectional study.

Empathy, as one of the fundamental principles of nursing professionalism, plays a pivotal role in the formation and advancement of the nursing team. Nursing interns, as a reserve force within the nursing team, are of significant importance in terms of their ability to empathize. This quality is not only directly related to the degree of harmony in the nurse-patient relationship and the enhancement of patient satisfaction, but also plays a pivotal role in the promotion of the quality of nursing services to a new level. The objective of this study was to gain a deeper understanding of the current state of nursing interns' empathic abilities. To this end, we sought to examine empathic performance under different profile models and to identify the key factors influencing these profile models. The study utilized 444 nursing interns from 11 tertiary general hospitals in Inner Mongolia as research subjects. The study employed a number of research tools, including demographic characteristics, the Jefferson Scale of Empathy, and the Professional Quality of Life Scale. A latent profile model of nursing interns' empathy ability was analyzed using Mplus 8.3. The test of variability of intergroup variables was performed using the chi-square test. Finally, the influencing factors of each profile model were analyzed by unordered multi-categorical logistic regression analysis. The overall level of empathy among nursing interns was found to be low, with 45% belonging to the humanistic care group, 43% exhibiting low empathy, and 12% demonstrating high empathy. The internship duration, empathy satisfaction, secondary traumatic stress, only child, place of birth, and satisfaction with nursing were identified as factors influencing the latent profiles of empathy in nursing interns (p < 0.05). There is considerable heterogeneity in nursing interns' ability to empathize. Consequently, nursing educators and administrators should direct greater attention to interns with lower empathy and develop targeted intervention strategies based on the influences of the different underlying profiles.

Estimating survival in data-driven phenotypes of mental health symptoms and peripheral biomarkers: A prospective study

BackgroundChronic psychological stress has widespread implications, including heightened mortality risk, mental and physical health conditions, and socioeconomic consequences. Stratified precision psychiatry shows promise in mitigating these effects by leveraging clinical heterogeneity to personalize interventions. However, little attention has been given to patient self-report. MethodsWe addressed this by combining stress-related self-report measures with peripheral biomarkers in a latent profile analysis and survival model. The latent profile models were estimated in a representative U.S. cohort (n = 1255; mean age = 57 years; 57% female) and cross-validated in Tokyo, Japan (n = 377; mean age = 55 years; 56% female). ResultsWe identified three distinct groups: “Good Mental Health”, “Poor Mental Health”, and “High Inflammation”. Compared to the “Good Mental Health” group, the “High Inflammation” and “Poor Mental Health” groups had an increased risk of mortality, but did not differ in mortality risk from each other. ConclusionsThis study emphasizes the role of patient self-report in stratified psychiatry.

A Spatial Semantic Feature Extraction Method for Urban Functional Zones Based on POIs

Accurately extracting semantic features of urban functional zones is crucial for understanding urban functional zone types and urban functional spatial structures. Points of interest provide comprehensive information for extracting the semantic features of urban functional zones. Many researchers have used topic models of natural language processing to extract the semantic features of urban functional zones from points of interest, but topic models cannot consider the spatial features of points of interest, which leads to the extracted semantic features of urban functional zones being incomplete. To consider the spatial features of points of interest when extracting semantic features of urban functional zones, this paper improves the Latent Dirichlet Allocation topic model and proposes a spatial semantic feature extraction method for urban functional zones based on points of interest. In the proposed method, an assumption (that points of interest belonging to the same semantic feature are spatially correlated) is introduced into the generation process of urban functional zones, and then, Gibbs sampling is combined to carry out the parameter inference process. We apply the proposed method to a simulated dataset and the point of interest dataset for Chaoyang District, Beijing, and compare the semantic features extracted by the proposed method with those extracted by the Latent Dirichlet Allocation. The results show that the proposed method sufficiently considers the spatial features of points of interest and has a higher capability of extracting the semantic features of urban functional zones than the Latent Dirichlet Allocation.