Parametric g-formula for Testing Time-Varying Causal Effects: What It Is, Why It Matters, and How to Implement It in Lavaan.

Psychologists leverage longitudinal designs to examine the causal effects of a focal predictor (i.e., treatment or exposure) over time. But causal inference of naturally observed time-varying treatments is complicated by treatment-dependent confounding in which earlier treatments affect confounders of later treatments. In this tutorial article, we introduce psychologists to an established solution to this problem from the causal inference literature: the parametric g-computation formula. We explain why the g-formula is effective at handling treatment-dependent confounding. We demonstrate that the parametric g-formula is conceptually intuitive, easy to implement, and well-suited for psychological research. We first clarify that the parametric g-formula essentially utilizes a series of statistical models to estimate the joint distribution of all post-treatment variables. These statistical models can be readily specified as standard multiple linear regression functions. We leverage this insight to implement the parametric g-formula using lavaan, a widely adopted R package for structural equation modeling. Moreover, we describe how the parametric g-formula may be used to estimate a marginal structural model whose causal parameters parsimoniously encode time-varying treatment effects. We hope this accessible introduction to the parametric g-formula will equip psychologists with an analytic tool to address their causal inquiries using longitudinal data.

Residential history, the electronic health record, and clinical medicine: A descriptive study in a lung transplant cohort.

Neighborhood level social determinants of health are commonly measured using a patient's most recent residential location. Not accounting for residential history, and therefore missing accumulated stressors from prior social vulnerabilities, could increase misclassification bias. We tested the hypothesis that the electronic health record could capture the residential history of lung transplant patients -a vulnerable population. After applying the Social Vulnerability Index (SVI) to individual residential histories, the most recent SVI equaled the first SVI in only 15.4% (58/374) of patients. There is a need for databases with residential histories to inform place-based determinants of health and applications to patient care.

The Assessment of the Association of Proton Pump Inhibitor Usage with Chronic Kidney Disease Progression through a Process Mining Approach.

Previous studies have suggested an association between Proton Pump Inhibitors (PPIs) and the progression of chronic kidney disease (CKD). This study aims to assess the association between PPI use and CKD progression by analysing estimated glomerular filtration rate (eGFR) trajectories using a process mining approach. We conducted a retrospective cohort study from 1 January 2006 to 31 December 2011, utilising data from the Stockholm Creatinine Measurements (SCREAM). New users of PPIs and H2 blockers (H2Bs) with CKD (eGFR < 60) were identified using a new-user and active-comparator design. Process mining discovery is a technique that discovers patterns and sequences in events over time, making it suitable for studying longitudinal eGFR trajectories. We used this technique to construct eGFR trajectory models for both PPI and H2B users. Our analysis indicated that PPI users exhibited more complex and rapidly declining eGFR trajectories compared to H2B users, with a 75% increased risk (adjusted hazard ratio [HR] 1.75, 95% confidence interval [CI] 1.49 to 2.06) of transitioning from moderate eGFR stage (G3) to more severe stages (G4 or G5). These findings suggest that PPI use is associated with an increased risk of CKD progression, demonstrating the utility of process mining for longitudinal analysis in epidemiology, leading to an improved understanding of disease progression.

Assessing tilavonemab efficacy in early Alzheimer's disease via longitudinal item response theory modeling.

INTRODUCTION: Alzheimer's disease (AD) is a neurodegenerative disorder characterized by declines in cognitive and functional severities. This research utilized the Clinical Dementia Rating (CDR) to assess the influence of tilavonemab on these deteriorations. METHODS: Longitudinal Item Response Theory (IRT) models were employed to analyze CDR domains in early-stage AD patients. Both unidimensional and multidimensional models were contrasted to elucidate the trajectories of cognitive and functional severities. RESULTS: We observed significant temporal increases in both cognitive and functional severities, with the cognitive severity deteriorating at a quicker rate. Tilavonemab did not demonstrate a statistically significant effect on the progression in either severity. Furthermore, a significant positive association was identified between the baselines and progression rates of both severities. DISCUSSION: While tilavonemab failed to mitigate impairment progression, our multidimensional IRT analysis illuminated the interconnected progression of cognitive and functional declines in AD, suggesting a comprehensive perspective on disease trajectories. Highlights: Utilized longitudinal Item Response Theory (IRT) models to analyze the Clinical Dementia Rating (CDR) domains in early-stage Alzheimer's disease (AD) patients, comparing unidimensional and multidimensional models.Observed significant temporal increases in both cognitive and functional severities, with cognitive severity deteriorating at a faster rate, while tilavonemab showed no statistically significant effect on either domain's progression.Found a significant positive association between the baseline severities and their progression rates, indicating interconnected progression patterns of cognitive and functional declines in AD.Introduced the application of multidimensional longitudinal IRT models to provide a comprehensive perspective on the trajectories of cognitive and functional severities in early AD, suggesting new avenues for future research including the inclusion of time-dependent random effects and data-driven IRT models.

A sparse factor model for clustering high-dimensional longitudinal data.

Recent advances in engineering technologies have enabled the collection of a large number of longitudinal features. This wealth of information presents unique opportunities for researchers to investigate the complex nature of diseases and uncover underlying disease mechanisms. However, analyzing such kind of data can be difficult due to its high dimensionality, heterogeneity and computational challenges. In this article, we propose a Bayesian nonparametric mixture model for clustering high-dimensional mixed-type (eg, continuous, discrete and categorical) longitudinal features. We employ a sparse factor model on the joint distribution of random effects and the key idea is to induce clustering at the latent factor level instead of the original data to escape the curse of dimensionality. The number of clusters is estimated through a Dirichlet process prior. An efficient Gibbs sampler is developed to estimate the posterior distribution of the model parameters. Analysis of real and simulated data is presented and discussed. Our study demonstrates that the proposed model serves as a useful analytical tool for clustering high-dimensional longitudinal data.

Ordinal Outcome State-Space Models for Intensive Longitudinal Data.

Intensive longitudinal (IL) data are increasingly prevalent in psychological science, coinciding with technological advancements that make it simple to deploy study designs such as daily diary and ecological momentary assessments. IL data are characterized by a rapid rate of data collection (1+ collections per day), over a period of time, allowing for the capture of the dynamics that underlie psychological and behavioral processes. One powerful framework for analyzing IL data is state-space modeling, where observed variables are considered measurements for underlying states (i.e., latent variables) that change together over time. However, state-space modeling has typically relied on continuous measurements, whereas psychological data often come in the form of ordinal measurements such as Likert scale items. In this manuscript, we develop a general estimation approach for state-space models with ordinal measurements, specifically focusing on a graded response model for Likert scale items. We evaluate the performance of our model and estimator against that of the commonly used "linear approximation" model, which treats ordinal measurements as though they are continuous. We find that our model resulted in unbiased estimates of the state dynamics, while the linear approximation resulted in strongly biased estimates of the state dynamics. Finally, we develop an approximate standard error, termed slice standard errors and show that these approximate standard errors are more liberal than true standard errors (i.e., smaller) at a consistent bias.

Machine learning predicts upper secondary education dropout as early as the end of primary school.

Education plays a pivotal role in alleviating poverty, driving economic growth, and empowering individuals, thereby significantly influencing societal and personal development. However, the persistent issue of school dropout poses a significant challenge, with its effects extending beyond the individual. While previous research has employed machine learning for dropout classification, these studies often suffer from a short-term focus, relying on data collected only a few years into the study period. This study expanded the modeling horizon by utilizing a 13-year longitudinal dataset, encompassing data from kindergarten to Grade 9. Our methodology incorporated a comprehensive range of parameters, including students' academic and cognitive skills, motivation, behavior, well-being, and officially recorded dropout data. The machine learning models developed in this study demonstrated notable classification ability, achieving a mean area under the curve (AUC) of 0.61 with data up to Grade 6 and an improved AUC of 0.65 with data up to Grade 9. Further data collection and independent correlational and causal analyses are crucial. In future iterations, such models may have the potential to proactively support educators' processes and existing protocols for identifying at-risk students, thereby potentially aiding in the reinvention of student retention and success strategies and ultimately contributing to improved educational outcomes.

A joint normal-ordinal (probit) model for ordinal and continuous longitudinal data.

In biomedical studies, continuous and ordinal longitudinal variables are frequently encountered. In many of these studies it is of interest to estimate the effect of one of these longitudinal variables on the other. Time-dependent covariates have, however, several limitations; they can, for example, not be included when the data is not collected at fixed intervals. The issues can be circumvented by implementing joint models, where two or more longitudinal variables are treated as a response and modeled with a correlated random effect. Next, by conditioning on these response(s), we can study the effect of one or more longitudinal variables on another. We propose a normal-ordinal(probit) joint model. First, we derive closed-form formulas to estimate the model-based correlations between the responses on their original scale. In addition, we derive the marginal model, where the interpretation is no longer conditional on the random effects. As a consequence, we can make predictions for a subvector of one response conditional on the other response and potentially a subvector of the history of the response. Next, we extend the approach to a high-dimensional case with more than two ordinal and/or continuous longitudinal variables. The methodology is applied to a case study where, among others, a longitudinal ordinal response is predicted with a longitudinal continuous variable.

Birth expectations, birth experiences and childbirth-related post-traumatic stress symptoms in mothers and birth companions: Dyadic investigation using response surface analysis.

OBJECTIVES: During the perinatal period, women and their birth companions form expectations about childbirth. We aimed to examine whether a mismatch between birth expectations and experiences predict childbirth-related post-traumatic stress symptoms (CB-PTSS) for mothers and birth companions. We also explored the influence of the mismatch between mothers' and birth companions' expectations/experiences on CB-PTSS. DESIGN: Dyadic longitudinal data from the Self-Hypnosis IntraPartum Trial. METHODS: Participants (n = 469 mothers; n = 358 birth companions) completed questionnaires at 27 and 36 weeks of gestation and 2 and 6 weeks post-partum. We used the measures of birth expectations (36 weeks gestation), birth experiences (2 weeks post-partum) and CB-PTSS (6 weeks post-partum). RESULTS: Correlations revealed that birth expectations were associated with experiences for both mothers and birth companions but were not consistently associated with CB-PTSS. Birth experiences related to CB-PTSS for both mothers and birth companions. The response surface analysis results showed no support for the effect of a mismatch between expectations and experiences on CB-PTSS in mothers or birth companions. Similarly, a mismatch between mothers' and birth companions' expectations or experiences was unrelated to CB-PTSS. CONCLUSIONS: Following previous literature, birth expectations were associated with experiences, and experiences were associated with CB-PTSS. By testing the effect of the match between birth experiences and expectations using an advanced statistical method, we found that experiences play a more substantial role than the match between experiences and expectations in CB-PTSS. The impact of birth experiences on CB-PTSS highlights the importance of respectful and supportive maternity care.

Person-specific dynamics between negative emotions and suicidal thoughts.

INTRODUCTION: Recent technology has enabled researchers to collect ecological momentary assessments (EMA) to examine within-person correlates of suicidal thoughts. Prior studies examined generalized temporal dynamics of emotions and suicidal thinking over brief periods, but it is not yet known how variable these processes are across people. METHOD: We use data EMA data delivered over two weeks with youth/young adults (N = 60) who reported past year self-injurious thoughts/behaviors. We used group iterative multiple model estimation (GIMME) to model group- and person-specific associations of negative emotions (i.e., fear, sadness, shame, guilt, and anger) and suicidal thoughts. RESULTS: 29 participants (48.33%) reported at least one instance of a suicidal thought and were included in GIMME models. In group level models, we consistently observed autoregressive effects for suicidal thoughts (e.g., earlier thoughts predicting later thoughts), although the magnitude and direction of this link varied from person-to-person. Among emotions, sadness was most frequently associated with contemporaneous suicidal thoughts, but this was evident for less than half of the sample, while other emotional correlates of suicidal thoughts broadly differed across people. No emotion variable was linked to future suicidal thoughts in >14% of the sample, CONCLUSIONS: Emotion-based correlates of suicidal thoughts are heterogeneous across people. Better understanding of the individual-level pathways maintaining suicidal thoughts/behaviors may lead to more effective, personalized interventions.

A longitudinal perspective to migrant health: Unpacking the immigrant health paradox in Germany.

Previous research finds that recent immigrants are healthier than the native-born, while more established immigrants exhibit worse health, suggesting a process of unhealthy assimilation. However, previous literature is mostly based on cross-sectional data or on longitudinal analyses similarly failing to disentangle individual-level variation from between-individual confounding. Moreover, previous longitudinal studies are often limited in their study of different health outcomes (few and mostly subjective health), populations (sometimes only elderly individuals), time periods (short panels) and geographical contexts (mostly Australia, Canada and USA). We address these limitations by comparing the health trajectories of adult immigrants and natives in Germany over extended periods, using data from years 2002-2021 of the German Socio-Economic Panel (SOEP), and investigating a wide range of health outcomes, including self-assessed physical and mental health measures, diagnosed illnesses, and health behaviors. We employ a longitudinal approach that stratifies immigrants by age at arrival, and compares them to natives of the same age. This allows us to estimate both Hierarchical Linear Models and more rigorous Fixed Effects models to further address confounding. Cross-sectionally, we confirm previous literature's findings: recent immigrants are healthier than natives and established immigrants. Longitudinally, we find support for the unhealthy assimilation hypothesis concerning subjective health and mental health, but not for the others health indicators or behaviors. We interpret these findings as possible evidence of immigrants' reduced access to timely health care and emphasize the need for greater longitudinal research investigating migrant gaps in various health outcomes.

Association of the Type of Public Pension With Mental Health Among South Korean Older Adults: Longitudinal Observational Study.

BACKGROUND: As income and health are closely related, retirement is considered undesirable for health. Many studies have shown the association between pension and health, but no research has considered the association between contribution-based public pensions or their types and health. OBJECTIVE: This study investigates the association between the type of contributory public pension and depressive symptoms among older adults. METHODS: We analyzed the data of 4541 older adults who participated in the South Korea Welfare Panel Study (2014-2020). Depressive symptoms were measured using the 11-item Center for Epidemiologic Studies Depression scale. Public pensions in South Korea are classified into specific corporate pensions and national pensions. For subgroup analyses, pensioners were categorized according to the amount of pension received and the proportion of public pension over gross income. Analyses using generalized estimating equations were conducted for longitudinal data. RESULTS: Individuals receiving public pension, regardless of the pension type, demonstrated significantly decreased depressive symptoms (national pension: ß=-.734; P<.001; specific corporate pension: ß=-.775; P=.02). For both pension types, the higher the amount of benefits, the lower were the depression scores. However, this association was absent for those who received the smaller amount among the specific corporate pensioners. In low-income households, the decrease in the depressive symptoms based on the amount of public pension benefits was greater (fourth quartile of national pension: ß=-1.472; P<.001; second and third quartiles of specific corporate pension: ß=-3.646; P<.001). CONCLUSIONS: Our study shows that contributory public pension is significantly associated with lower depressive symptoms, and this association is prominent in low-income households. Thus, contributory public pensions may be good income sources for improving the mental health of older adults after retirement.

1H-nuclear magnetic resonance analysis reveals dynamic changes in the metabolic profile of patients with severe burns.

Background: Severe burn injury causes a hypermetabolic response, resulting in muscle protein catabolism and multiple organ damage syndrome. However, this response has not yet been continuously characterized by metabolomics in patients. This study aims to quantify temporal changes in the metabolic processes of patients with severe burns. Methods: We employed 1H-nuclear magnetic resonance (NMR) spectroscopy to scrutinize metabolic alterations during the initial 35 days following burn injury in a cohort of 17 adult patients with severe burns, with 10 healthy individuals included as controls. Plasma specimens were collected from patients on postburn days 1, 3, 7, 14, 21, 28 and 35. After performing multivariate statistical analysis, repeated-measures analysis of variance and time-series analysis, we quantified changes in metabolite concentrations. Results: Among the 36 metabolites quantified across 119 samples from burn patients, branched-chain amino acids, glutamate, glycine, glucose, pyruvate, lactate, trimethylamine N-oxide and others exhibited obvious temporal variations in concentration. Notably, these metabolites could be categorized into three clusters based on their temporal characteristics. The initial response to injury was characterized by changes in lactate and amino acids, while later changes were driven by an increase in fatty acid catabolism and microbial metabolism, leading to the accumulation of ketone bodies and microbial metabolites. Conclusions: Metabolomics techniques utilizing NMR have the potential to monitor the intricate processes of metabolism in patients with severe burns. This study confirmed that the third day after burn injury serves as the boundary between the ebb phase and the flow phase. Furthermore, identification of three distinct temporal patterns of metabolites revealed the intrinsic temporal relationships between these metabolites, providing clinical data for optimizing therapeutic strategies.

The relations among prosocial behavior, hedonic, and eudaimonic well-being in everyday life.

INTRODUCTION: Existing research highlights the significance of prosocial behavior (voluntary, intentional behavior that results in benefits for another) to people's well-being. Yet, the extent to which this expected positive relation operates at the within-person level (e.g., is more prosocial behavior than usual related to a higher than usual level of well-being?) while taking into account stable interindividual differences, remains a research question that deserves further investigation. In this study, we aimed to explore the relations between prosocial behavior and hedonic (HWB; subjective assessment of life satisfaction and happiness) and eudaimonic (EWB; actualization of human potential in alignment with personal goals, including concepts like meaning in life and closeness to others) well-being in daily life. METHOD: Using ecological momentary assessment for 4 weeks, data were collected from two British samples, comprising 82 adolescents and 166 adults. RESULTS: Dynamic Structural Equation Modeling revealed a positive relations between prosocial behavior and HWB/EWB at both between and within-person levels across the samples. CONCLUSION: In summary, these findings further support the positive link between prosocial behavior and well-being in everyday life. Notably, this association was consistent across different age groups (adolescent and adults) at both between and within-person levels.

Life course predictors of child emotional distress during the COVID-19 pandemic: Findings from a prospective intergenerational cohort study.

BACKGROUND: We examine precursors of child emotional distress during the COVID-19 pandemic in a prospective intergenerational Australian cohort study. METHODS: Parents (N = 549, 60% mothers) of 934 1-9-year-old children completed a COVID-19 specific module in 2020 and/or 2021. Decades prior, a broad range of individual, relational and contextual factors were assessed during parents' own childhood, adolescence and young adulthood (7-8 to 27-28 years old; 1990-2010) and again when their children were 1 year old (2012-2019). RESULTS: After controlling for pre-pandemic socio-emotional behaviour problems, COVID-19 child emotional distress was associated with a range of pre-pandemic parental life course factors including internalising difficulties, lower conscientiousness, social skills problems, poorer relational health and lower trust and tolerance. Additionally, in the postpartum period, pre-pandemic parental internalising difficulties, lower parental warmth, lower cooperation and fewer behavioural competencies predicted child COVID-19 emotional distress. CONCLUSIONS: Findings highlight the importance of taking a larger, intergenerational perspective to better equip young populations for future adversities. This involves not only investing in child, adolescent, and young adult emotional and relational health, but also in parents raising young families.

Bayesian variable selection in joint modeling of longitudinal data and interval-censored failure time data.

Joint modeling of longitudinal data and survival data has gained great attention in the last two decades. However, most of the existing studies have focused on right-censored survival data. In this article, we study joint analysis of longitudinal data and interval-censored survival data and conduct Bayesian variable selection in this framework. A new joint model is proposed with a shared frailty to characterize the dependence between the two types of responses, where the longitudinal response is modeled with a semiparametric linear mixed-effects submodel and the survival time is modeled by a semiparametric normal fraility probit sub-model. Several Bayesian variable selection approaches are developed by adopting Bayesian Lasso, adaptive Lasso, and spike-and-slab priors in order to simultaneously select significant covariates and estimate their effects on the two types of responses. Efficient Gibbs samplers are proposed with all unknown parameters and latent variables being sampled directly from well recognized full conditional distributions. Our simulation study shows that these methods perform well in both variable selection and parameter estimation. A real-life data application to joint analysis of blood cholesterol level and hypertension is provided as an illustration.

Predictors of male loneliness across life stages: an Australian study of longitudinal data.

BACKGROUND: Despite growing recognition of loneliness as a global public health concern, research on its occurrence and precipitants among men across different life stages remains limited and inconclusive. This study aims to address this gap by investigating the prevalence and predictors of loneliness among a large, representative data set of Australian adult men. METHODS: The study used longitudinal data from waves 2-21 of the Household, Income and Labour Dynamics in Australia (HILDA) Survey, including men aged 15-98. Estimating linear fixed effects regressions that account for unobserved time-invariant individual heterogeneity, a single-item measure of loneliness was regressed on a set of selected explanatory variables over different parts of the life course. RESULTS: Increased social isolation, romantic partnership dissolution, having a long-term disability, and stronger beliefs that the man, rather than the woman, should be the breadwinner of the household, are associated with greater loneliness. Frequent social connection, having a romantic partner, and high neighbourhood satisfaction are protective against loneliness. The findings also reveal several differences in the predictors of loneliness over the life course. Job security is especially important for younger men, whereas for older men volunteering and less conservative gender role attitudes are important factors that can decrease loneliness. CONCLUSIONS: The results emphasise the need to consider age-specific factors and societal expectations in understanding and addressing loneliness amongst men. Additionally, the findings underscore the importance of raising awareness about the impact of societal norms and expectations on men's mental health. The results offer valuable insights for policymakers, healthcare providers, and researchers to develop effective strategies and support systems to combat loneliness and promote well-being among men.

Longitudinal effects of sexual minority stigma on depressive symptoms among young men who have sex with men: Distinguish between and within person effects.

BACKGROUND: Young men who have sex with men (YMSM) may experience high levels of sexual minority stigma (SMS) and depressive symptoms (DS) over the world and in China. However, there is a lack of studies investigating the longitudinal effects of SMS on DS of YMSM, especially focusing on YMSM and separating the between-person and within-person effects. This study aimed to fill the said gaps. METHODS: Study data were derived from a prospective cohort of 349 YMSM from central China (Wuhan, Changsha, Nanchang), the baseline survey was started in 2017 with one follow-up visit every year. SMS and DS were measured three times using valid and reliable instruments. The cross-lagged panel model (CLPM) and the random intercept CLPM (RI-CLPM) were used to examine the between-person and within-person concurrent and lagged effects, respectively. RESULTS: Findings of CLPM revealed bidirectional associations between SMS and DS over time. RI-CLPM suggested that at the between-person level, SMS was significantly associated with DS, echoing the results of CLPM. However, this reciprocal relationship has not been found at the within-person level. CONCLUSION: The associations between SMS and DS among YMSM at the population level is more significant than that at the individual level. We suggest that interventions should be against the adverse effects of cultural marginalization and systemic change the social concepts to reduce the amount of SMS in society.

Predicting Injuries in Elite Female Football Players With Global-Positioning-System and Multiomics Data.

PURPOSE: Injury prevention is a crucial aspect of sports, particularly in high-performance settings such as elite female football. This study aimed to develop an injury prediction model that incorporates clinical, Global-Positioning-System (GPS), and multiomics (genomics and metabolomics) data to better understand the factors associated with injury in elite female football players. METHODS: We designed a prospective cohort study over 2 seasons (2019-20 and 2021-22) of noncontact injuries in 24 elite female players in the Spanish Premiership competition. We used GPS data to determine external workload, genomic data to capture genetic susceptibility, and metabolomic data to measure internal workload. RESULTS: Forty noncontact injuries were recorded, the most frequent of which were muscle (63%) and ligament (20%) injuries. The baseline risk model included fat mass and the random effect of the player. Six genetic polymorphisms located at the DCN, ADAMTS5, ESRRB, VEGFA, and MMP1 genes were associated with injuries after adjusting for player load (P < .05). The genetic score created with these 6 variants determined groups of players with different profile risks (P = 3.1 × 10-4). Three metabolites (alanine, serotonin, and 5-hydroxy-tryptophan) correlated with injuries. The model comprising baseline variables, genetic score, and player load showed the best prediction capacity (C-index: .74). CONCLUSIONS: Our model could allow efficient, personalized interventions based on an athlete's vulnerability. However, we emphasize the necessity for further research in female athletes with an emphasis on validation studies involving other teams and individuals. By expanding the scope of our research and incorporating diverse populations, we can bolster the generalizability and robustness of our proposed model.