Within-subject Correlation Research Articles

Latent Gaussian copula models for longitudinal binary data

Longitudinal binary data arise commonly in a variety of fields including public health, biomedicine, finance, agriculture and social science, among many others. In longitudinal binary studies, the aims are to assess the association of longitudinal binary response with certain covariates of interest, and to quantify the within-subject correlations for longitudinal binary responses. In the literature, various methods were developed to model longitudinal binary data but little work was done to account for the fact that the correlation coefficients of correlated binary responses have the so-called Fréchet–Hoeffding bounds. Ignoring this fact can lead to incorrect statistical inferences for longitudinal binary data. In this paper, based on latent Gaussian copula a new statistical modeling method is proposed to model the mean and within-subject correlation structures, simultaneously, for longitudinal binary data. Specifically, the mean structure is modeled by a semiparametric regression model, and the within-subject correlation coefficients are modeled through introducing a latent Gaussian copula model with certain latent correlation structures characterized by some parameters. Generalized estimating equations are then proposed to estimate the parameters in the mean and latent correlation structures, and consistency and asymptotic normality of the resulting parameter estimators are established. The proposed model and method ensure that the estimated correlation coefficients must satisfy the Fréchet–Hoeffding bounds for longitudinal binary data. Simulation studies show that the proposed method has a stable numerical performance. A practical data set is analyzed using the proposed method for illustration.

Longitudinal Association Between Angina Pectoris and Quality of Life

Angina is a common symptom in patients with coronary artery disease (CAD); however, its impact on patients' quality of life over time is not well understood. We sought to determine the longitudinal association of angina frequency with quality of life and functional status over a 5-year period. We used data from the Heart and Soul Study, a prospective cohort study of 1,023 outpatients with stable CAD. Participants completed the Seattle Angina Questionnaire (SAQ) at baseline and annually for 5 years. We evaluated the population effect of angina frequency on disease-specific quality of life (SAQ Disease Perception), physical function (SAQ Physical Limitation), perceived overall health, and overall quality of life, with adjusted models. We evaluated these associations within the same year and with a time-lagged association between angina and quality of life reported 1 year later. Generalized estimating equation models were used to account for repeated measures and within-subject correlation of responses. Over 5years of follow-up, patients with daily or weekly angina symptoms had lower quality of life scores (52 vs 89, p <0.001) and greater physical limitation (61 vs 86, p <0.001) after adjustment. Compared with patients with daily or weekly angina symptoms, those with no angina symptoms had 2-fold greater odds of better quality of life (odds ratio 2.39, 95% confidence interval 1.76 to 3.25) and 5-fold greater odds of better perceived overall health (odds ratio 5.45, 95% confidence interval 3.85 to 7.73). In conclusion, angina frequency is strongly associated with quality of life and physical function in patients with CAD. Even after modeling to adjust for both clinical risk factors and repeated measures within subjects, we found that less frequent angina symptoms were associated with better quality of life.

Shared parameter and copula models for analysis of semicontinuous longitudinal data with nonrandom dropout and informative censoring.

Analysis of longitudinal semicontinuous data characterized by subjects' attrition triggered by nonrandom dropout is complex and requires accounting for the within-subject correlation, and modeling of the dropout process. While methods that address the within-subject correlation and missing data are available, approaches that incorporate the nonrandom dropout, also referred to informative right censoring, in the modeling step are scarce due to the computational intensity and possible intractable integration needed for its implementation. Appreciating the complexity of this problem and the need for a new methodology that is feasible for implementation, we propose to extend a framework of likelihood-based marginalized two-part models to account for informative right censoring. The censoring process is modeled using two approaches: (1) Poisson censoring for the count of visits before dropout and (2) survival time to dropout. Novel consideration was given to the proposed joint modeling approaches for the semicontinuous and censoring components of the likelihood function which included (1) shared parameter, and (2) Clayton copula. The cross-part and within-part correlations were accounted for through a complex random effect structure that models correlated random intercepts and slopes. Feasibility of implementation, and accuracy of these approaches were investigated using extensive simulation studies and clinical application.

Hemodynamic Responses to Crystalloid and Colloid Fluid Boluses during Noncardiac Surgery.

Colloids are thought to sustain blood pressure and cardiac index better than crystalloids. However, the relative effects of intraoperative hydroxyethyl starch and crystalloid administration on the cardiac index and blood pressure remain unclear. This study therefore tested in this subanalysis of a previously published large randomized trial the hypothesis that intraoperative goal-directed colloid administration increases the cardiac index more than goal-directed crystalloid administration. Further, the effects of crystalloid and colloid boluses on blood pressure were evaluated. This planned subanalysis of a previous trial analyzed data from 973 patients, of whom 480 were randomized to colloids and 493 were randomized to crystalloids. Fluid administration was guided by esophageal Doppler. The primary outcome was the time-weighted average cardiac index during surgery between the colloid and crystalloid group. The secondary outcomes were the cardiac index just after bolus administration, time elapsed between boluses, and the average real variability during surgery. The study recorded cardiac index, corrected flow time, and blood pressure at 10-min intervals, as well as before and after each bolus. Time-weighted average of cardiac index over the duration of anesthesia was only slightly greater in patients given colloid than crystalloid, with the difference being just 0.20 l · min-1 · m-2 (95% CI, 0.11 to 0.29; P < 0.001). However, the hazard for needing additional boluses was lower after colloid administration (hazard ratio [95% CI], 0.60 [0.55 to 0.66]; P < 0.001) in a frailty time-to-event model accounting for within-subject correlation. The median [quartiles] number of boluses per patient was 4 [2, 6] for colloids and 6 [3, 8] for crystalloids, with a median difference (95% CI) of -1.5 (-2 to -1; P < 0.001). The average real mean arterial pressure variability did not differ significantly between the groups (difference in means [95% CI] of -0.03 (-0.07 to 0.02) mmHg, P = 0.229). There were not clinically meaningful differences in the cardiac index or mean pressure variability in patients given goal-directed colloid and crystalloids. As might be expected from longer intravascular dwell time, the interval between boluses was longer with colloids. However, on a case basis, the number of boluses differed only slightly. Colloids do not appear to provide substantial hemodynamic benefit.

Thirteen-year follow-up of long-term treated psychotic disorder: personality aspects

Objective Personality is an aspect that can affect the symptoms and social function in individuals with psychotic disorders. Several studies have investigated personality in schizophrenia and other long-term psychotic disorders. No study has examined the stability of personality traits exceeding five years in patients with schizophrenia and related disorders. The aim of this study was to investigate the stability of personality traits over a 13-year period among patients with schizophrenia and related disorders and healthy individuals and to evaluate case-control differences. Methods At three occasions during a 13-year period patients with schizophrenia and related disorders (n = 28) and healthy individuals (n = 57) completed Swedish universities Scales of Personality (SSP). Mean-level change and case-control differences were investigated for all the individuals using within- and between-subject analyses, respectively. Analyses were performed on three occasions for all 13 subscales and the three overall factors of SSP. Also, correlations, means, and SDs were calculated. Results Tests of within-subject correlations showed differences in two subscales: Lack of Assertiveness, which were influenced by age, and Physical Trait Aggression, where patients’ ratings were stable, whereas controls rated themselves less aggressive at a higher age. Between-subjects correlations showed differences regarding diagnosis, time, age, gender, or age × gender in nine of the 13 subscales as well as in factor Neuroticism. Conclusion Long-term follow-up showed generally high stability of personality traits measured with SSP. Between-subject analyses over the 13 years showed that patients differed compared to controls for the SSP factor Neuroticism as well as the subscale Detachment, which is in accordance with previous studies within this population.

Comparing two statistical models for low boom dose-response relationships with correlated responses

This study compares two statistical modeling approaches to construct summary dose-response curves for low boom community noise surveys. NASA field survey data were used that consist of multiple responses from a survey participant. These data require an approach that accounts for the correlation among repeated annoyance observations from the same participant. A multilevel model accounts for the correlation by allowing estimated parameters to vary with each survey participant. On the other hand, a population average model utilizes generalized estimating equations and accounts for the correlation via a user-specified within-subject correlation structure. A visual comparison of the dose-response curves for these two methods reveals similar results. When comparing specific points along the summary curves, the multilevel model yields more precise confidence bounds than the population average model. The similarity between the summary curves derived from each model lends validity to both approaches for approximating a population representative summary curve.

Gender-specific associations between psychological distress and injecting risk behaviours among people who inject drugs in Montreal, Canada.

Psychological distress is common among people who inject drugs (PWID) and may be associated with HCV-related risk behaviours. Previous studies have documented increased vulnerability to both psychological distress and HCV infection among female relative to male PWID. It is, however, unclear whether behavioural responses to distress differ by gender. This study estimated gender-specific associations between psychological distress and i) binge drug injection, and ii) sharing of injection materials. Data were drawn from HEPCO, a longitudinal cohort study involving three-monthly interviews with active PWID in Montreal, Canada. Past-month psychological distress was assessed with the Kessler (K10) scale, categorized for descriptive analyses as minimal (score 10-15), moderate (16-21), high (22-29), or severe (30-50). Binge was defined as injecting large quantities of drugs until participants could no longer continue (past 3 months). Sharing was defined as injection with previously-used needles or equipment (past 3 months). Generalized additive models were fit to estimate smooth, nonlinear associations between K10 scores and risk behaviours, by gender. Models were adjusted for known determinants of drug-related harms and included random intercepts to model within-subject correlation. 805 individuals (82% male) provided 8158 observations (2011-2020). High to severe levels of distress were common and more frequent among women (55% vs 37%). Among men, the odds of binge and sharing monotonically and non-linearly increased with increasing scores of psychological distress. Associations for binge among women were attenuated relative to men but nevertheless increased with distress, albeit in a linear fashion. Sharing was not associated with distress among women. Psychological distress was differentially associated with injecting risk behaviours among men and women who inject drugs. Assessment of distress may provide novel prevention opportunities for select PWID. Further investigation into gender differences is warranted to inform development and tailoring of interventions.

Penalized joint generalized estimating equations for longitudinal binary data.

In statistical research, variable selection and feature extraction are a typical issue. Variable selection in linear models has been fully developed, while it has received relatively little attention for longitudinal data. Since a longitudinal study involves within-subject correlations, the likelihood function of discrete longitudinal responses generally cannot be expressed in analytically closed form, and standard variable selection methods cannot be directly applied. As an alternative, the penalized generalized estimating equation (PGEE) is helpful but very likely results in incorrect variable selection if the working correlation matrix is misspecified. In many circumstances, the within-subject correlations are of interest and need to be modeled together with the mean. For longitudinal binary data, it becomes more challenging because the within-subject correlation coefficients have the so-called Fréchet-Hoeffding upper bound. In this paper, we proposed smoothly clipped absolute deviation (SCAD)-based and least absolute shrinkage and selection operator (LASSO)-based penalized joint generalized estimating equation (PJGEE) methods to simultaneously model the mean and correlations for longitudinal binary data, together with variable selection in the mean model. The estimated correlation coefficients satisfy the upper bound constraints. Simulation studies under different scenarios are made to assess the performance of the proposed method. Compared to existing PGEE methods that specify a working correlation matrix for longitudinal binary data, the proposed PJGEE method works much better in terms of variable selection consistency and parameter estimation accuracy. A real data set on Clinical Global Impression is analyzed forillustration.

Predicting Driver Fatigue in Monotonous Automated Driving with Explanation using GPBoost and SHAP

ABSTRACT Research indicates that monotonous automated driving increases the incidence of fatigued driving. Although many prediction models based on advanced machine learning techniques were proposed to monitor driver fatigue, especially in manual driving, little is known about how these black-box machine learning models work. In this paper, we proposed a combination of Gaussian Process Boosting (GPBoost) and SHapley Additive exPlanations (SHAP) to predict driver fatigue with explanations. First, in order to obtain the ground truth of driver fatigue, we used PERCLOS (percentage of eyelid closure over the pupil over time) between 0 and 100 as the response variable. Second, we built a driver fatigue regression model using both physiological and behavioral measures with GPBoost that was able to address the within-subjects correlations. This model outperformed other selected machine learning models with root-mean-squared error (RMSE) = 2.965, mean absolute error (MAE) = 1.407, and adjusted . Third, we employed SHAP to identify the most important predictor variables and uncovered the black-box GPBoost model by showing the main effects of the most important predictor variables globally and explaining individual predictions locally. Such an explainable driver fatigue prediction model offered insights into how to intervene in automated driving when necessary, such as during the takeover transition period from automated driving to manual driving.

Modelling menstrual cycle length in athletes using state-space models

The ability to predict an individual’s menstrual cycle length to a high degree of precision could help female athletes to track their period and tailor their training and nutrition correspondingly. Such individualisation is possible and necessary, given the known inter-individual variation in cycle length. To achieve this, a hybrid predictive model was built using data on 16,524 cycles collected from a sample of 2125 women (mean age 34.38 years, range 18.00–47.10, number of menstrual cycles ranging from 4 to 53). A mixed-effect state-space model was fitted to capture the within-subject temporal correlation, incorporating a Bayesian approach for process forecasting to predict the duration (in days) of the next menstrual cycle. The modelling procedure was split into three steps (1) a time trend component using a random walk with an overdispersion parameter, (2) an autocorrelation component using an autoregressive moving-average model, and (3) a linear predictor to account for covariates (e.g. injury, stomach cramps, training intensity). The inclusion of an overdispersion parameter suggested that 26.36%[23.68%,29.17%] of cycles in the sample were overdispersed. The random walk standard deviation for a non-overdispersed cycle is 27.41 pm 1.05 [1.00, 1.09] days while under an overdispersed cycle, the menstrual cycle variance increase in 4.78 [4.57, 5.00] days. To assess the performance and prediction accuracy of the model, each woman’s last observation was used as test data. The root mean square error (RMSE), concordance correlation coefficient and Pearson correlation coefficient (r) between the observed and predicted values were calculated. The model had an RMSE of 1.6412 days, a precision of 0.7361 and overall accuracy of 0.9871. In conclusion, the hybrid model presented here is a helpful approach for predicting menstrual cycle length, which in turn can be used to support female athlete wellness.

Validity of electrodermal activity-based measures of sympathetic nervous system activity from a wrist-worn device

Measuring electrodermal activity (EDA) on the wrist with the use of dry electrodes is a promising method to help identify person-specific stressors during prolonged recordings in daily life. While the feasibility of this method has been demonstrated, detailed testing of validity of such ambulatory EDA is scarce. In a controlled laboratory study, we examine SCL and ns.SCR derived from wrist-based dry electrodes (Philips DTI) and palm-based wet electrodes (VU-AMS) in 112 healthy adults (57% females, mean age = 22.3, SD = 3.4) across 26 different conditions involving mental stressors or physical activities. Changes in these EDA measures were compared to changes in the Pre-ejection period (PEP) and stressor-induced changes in affect. Absolute SCL and ns.SCR frequency were lower at the wrist compared to the palm. Wrist-based ns.SCR and palm-based ns.SCR and SCL responded directionally consistent with our experimental manipulation of sympathetic nervous system (SNS) activity. Average within-subject correlations between palm-based and wrist-based EDA were significant but modest (r SCL = 0.31; r ns.SCR = 0.42). Changes in ns.SCR frequency at the palm (r = −0.44) and the wrist (r = −0.36) were correlated with changes in PEP. Both palm-based and wrist based EDA predicted changes in affect (6.5%–14.5%). Our data suggest that wrist-based ns.SCR frequency is a useful addition to the psychophysiologist's toolkit, at least for epidemiology-sized ambulatory studies of changes in sympathetic activity during daily life.

On the design of early-phase Alzheimer's disease clinical trials with cerebrospinal fluid tau outcomes.

The focus of Alzheimer's disease studies has shifted to earlier disease stages, including mild cognitive impairment. Biomarker inclusion criteria are often incorporated into mild cognitive impairment clinical trials to identify individuals with "prodromal Alzheimer's disease" to ensure appropriate drug targets and enrich for participants likely to develop Alzheimer's disease dementia. The use of these eligibility criteria may affect study power. We investigated outcome variability and study power in the setting of proof-of-concept prodromal Alzheimer's disease trials that incorporate cerebrospinal fluid levels of total tau (t-tau) and phosphorylated (p-tau) as primary outcomes and how differing biomarker inclusion criteria affect power. We used data from the Alzheimer's Disease Neuroimaging Initiative to model trial scenarios and to estimate the variance and within-subject correlation of total and phosphorylated tau. These estimates were then used to investigate the differences in study power for trials considering these two surrogate outcomes. Patient characteristics were similar for all eligibility criteria. The lowest outcome variance and highest within-subject correlation were obtained when phosphorylated tau was used as an eligibility criterion, compared to amyloid beta or total tau, regardless of whether total tau or phosphorylated tau were used as primary outcomes. Power increased when eligibility criteria were broadened to allow for enrollment of subjects with either low amyloid beta or high phosphorylated tau. Specific biomarker inclusion criteria may impact statistical power in trials using total tau or phosphorylated tau as the primary outcome. In concert with other important considerations such as treatment target and population of clinical interest, these results may have implications to the integrity and efficiency of prodromal Alzheimer's disease trial designs.

Circadian periodicity affects the type of ventricular arrhythmias and efficacy of implantable defibrillator therapies.

Factors influencing malignant arrhythmia onset are not fully understood. We explored the circadian periodicity of ventricular arrhythmias (VAs) in patients with implantable cardioverter and cardiac resynchronization defibrillators (ICD/CRT-D). Time, morphology (monomorphic/polymorphic), and mode of termination (anti-tachycardia pacing [ATP] or shock) of VAs stored in a database of remote monitoring data were adjudicated. Episodes were grouped in six 4-h timeslots from 00:00 to 24:00. Circadian distributions and adjusted marginal odds ratios (ORs), with 95% confidence interval (CI), were analyzed using mixed-effect models and logit generalized estimating equations, respectively, to account for within-subject correlation of multiple episodes. Among 1303 VA episodes from 446 patients (63% ICD and 37% CRT-D), 120 (9%) self-extinguished, and 842 (65%) were terminated by ATP, 343 (26%) by shock. VAs clustered from 08:00 to 16:00 with 44% of episodes, as compared with 22% from 00:00 to 08:00 (p < .001) and 34% from 16:00 to 24:00 (p = .005). Episodes were more likely to be polymorphic at night with an adjusted marginal OR of 1.66 (CI, 1.15-2.40; p = .007) at 00:00-04:00 versus other timeslots. Episodes were less likely to be terminated by ATP in the 00:00-04:00 (success-to-failure ratio, 0.67; CI, 0.46-0.98; p = .039) and 08:00-12:00 (0.70; CI, 0.51-0.96; p = .02) timeslots, and most likely to be terminated by ATP between 12:00 and 16:00 (success-to-failure ratio 1.42; CI, 1.06-1.91; p = .02). VAs did not distribute uniformly over the 24 h, with a majority of episodes occurring from 08:00 to 16:00. Nocturnal episodes were more likely to be polymorphic. The efficacy of ATP depended on the time of delivery.

Physical activity, patient-reported symptoms, and clinical events: Insights into postprocedural recovery from personal digital devices

BackgroundPersonal digital devices may offer insights into patient recovery and an approach for remote monitoring after procedures.ObjectiveTo examine associations between activity measured using personal digital devices, patient-reported outcome measures (PROMs), and clinical events among patients after catheter ablation for atrial fibrillation (AF) or bariatric surgery.MethodsWe aggregated personal digital device, PROM, and electronic health record data in a study conducted at 2 health systems. We used Fitbit devices for step count assessments, KardiaMobile for cardiac rhythm assessments, and PROMs for pain and palpitations over 5 weeks.ResultsAmong 59 patients, 30 underwent AF ablation and 29 bariatric surgery. Thirty-six patients (63%) reported pain. There was no difference in median [interquartile range] daily steps between patients with and those without pain (4419 [3286–7041] vs 3498 [2609–5888]; P = .23). Among AF ablation patients, 21 (70%) reported palpitations. Median daily steps were lower among those with palpitations than among those without (4668 [3021–6116] vs 8040 [6853–10,394]; P = .03). When accounting for within-subject correlation, recordings of AF were associated with a significant mean decrease in median daily steps (–351; 95% confidence interval –524 to –177; P <.01). Patients who received a new antiarrhythmic drug prescription had AF recorded in a median of 5 [5–5] of 5 total weeks, whereas patients who did not receive a new antiarrhythmic recorded AF in a median of 1 [0–3] week (P = .02).ConclusionPersonal digital device and PROM data can provide insight into postprocedural recovery outside of usual clinical settings and may inform follow-up and clinical decision-making. (ClinicalTrials.gov Identifier: NCT03436082)

Statistical Modeling of Students’ Academic Performances: A Longitudinal Study

Within students’ attrition studies, it is necessary to assess the longitudinal evolution of students within a given course of study, from enrolment to exit from the university through degree completion and academic dropout. Here, the student's academic progress is monitored through the number of courses failed each semester enrolled. The students’ failure rate and academic behavior typically provide significant insight into students’ exit outcomes from University programs. These programs usually have a maximum time frame required to complete the course. A likelihood-based approach is discussed that conditions on the exit outcome and random effects in adjusting within-subject correlation of longitudinal measurements. Ignoring the number of courses enrolled by a student may produce inadequate results on the actual failure rates. Conditioned on the exit outcomes of the student, we find out that factors such as financial aid, matriculation points, students’ race and course type registered, and gender are distinguishing factors that affect students’ academic performances, for completers and dropouts. Also, being in university-type accommodation (that often have added services such as transportation and internet connections) does not seem to significantly affect the failure rate within both groups of students. In addition, an increase in matriculation points significantly reduces the failure rate independent of the Quintile school of the student. Several count models such as mixed Poisson, mixed Zero Inflation Poisson, mixed Negative Binomial, and mixed Hurdle Poisson models are fitted and compared. In particular, the mixed Poisson model provides a better fit based on the Bayesian Information Criterion and residues analysis.

Increased extravascular lung water index (EVLWI) reflects rapid non-cardiogenic oedema and mortality in COVID-19 associated ARDS

Nearly 5% of patients suffering from COVID-19 develop acute respiratory distress syndrome (ARDS). Extravascular lung water index (EVLWI) is a marker of pulmonary oedema which is associated with mortality in ARDS. In this study, we evaluate whether EVLWI is higher in patients with COVID-19 associated ARDS as compared to COVID-19 negative, ventilated patients with ARDS and whether EVLWI has the potential to monitor disease progression. EVLWI and cardiac function were monitored by transpulmonary thermodilution in 25 patients with COVID-19 ARDS subsequent to intubation and compared to a control group of 49 non-COVID-19 ARDS patients. At intubation, EVLWI was noticeably elevated and significantly higher in COVID-19 patients than in the control group (17 (11–38) vs. 11 (6–26) mL/kg; p < 0.001). High pulmonary vascular permeability index values (2.9 (1.0–5.2) versus 1.9 (1.0–5.2); p = 0.003) suggested a non-cardiogenic pulmonary oedema. By contrast, the cardiac parameters SVI, GEF and GEDVI were comparable in both cohorts. High EVLWI values were associated with viral persistence, prolonged intensive care treatment and in-hospital mortality (23.2 ± 6.7% vs. 30.3 ± 6.0%, p = 0.025). Also, EVLWI showed a significant between-subjects (r = − 0.60; p = 0.001) and within-subjects correlation (r = − 0.27; p = 0.028) to Horowitz index. Compared to non COVID-19 ARDS, COVID-19 results in markedly elevated EVLWI-values in patients with ARDS. High EVLWI reflects a non-cardiogenic pulmonary oedema in COVID-19 ARDS and could serve as parameter to monitor ARDS progression on ICU.

Long-Term Trends in Ultra-Processed Food Consumption by Cardiometabolic Disease Status in the Framingham Offspring Study

Abstract Objectives Ultra-processed foods currently provide the majority of calories in the U.S. and are closely linked with cardiometabolic diseases. Yet, little is known regarding consumption trends among individuals with and without cardiometabolic diseases. The current analyses examined longitudinal trends in diet processing level from 1991–2008 within the Framingham Heart Study Offspring Cohort by disease status. Methods Analyses were conducted using dietary data serially collected by a food frequency questionnaire at study exams every 4 years between 1991–2008. The analytical sample included 2893 adults with valid dietary data for ≥3 exams (mean age: 54.0 y). Based on the NOVA framework, foods were classified as unprocessed/minimally processed foods, processed culinary ingredients (salt, sugar, fats and oils), culinary preparations (homemade baked goods and dishes), processed foods and ultra-processed foods. Consumption of each processing level (servings/d) were examined by calculating means, standard deviations and p-values for trends over the 4 exams. Mixed effects models with subject-specific random intercepts were used to account for within-subject correlation. Trends were stratified by prevalent cardiometabolic disease (CVD, diabetes, hypertension). Results Among individuals with cardiometabolic diseases, over 17 years of follow-up, consumption of ultra-processed foods (7.9–6.1 servings/d) and minimally processed foods (11.2–10.6 servings/d) decreased (p-trend &amp;lt; 0.001). Intakes of processed foods, culinary ingredients and culinary preparations changed minimally during the study period. Trends were similar among individuals without cardiometabolic diseases (ultra-processed foods 7.3 to 5.8 servings/d) and minimally processed foods (11.3–11.0 servings/d) (p-trend &amp;lt; 0.001). Conclusions Consumption of ultra-processed foods, a putative cardiovascular risk factor, remained high throughout the 1990s and 2000s, among aging adults, regardless of the cardiometabolic disease status. Yet, processing level has not been addressed in federal or clinical dietary guidelines. Future investigations should determine if dietary recommendations distinguishing ultra-processed from moderately and minimally processed foods facilitate healthier food choices. Funding Sources None.

Awareness of sleepiness: Temporal dynamics of subjective and objective sleepiness.

We systematically examined the temporal relationships between subjective sleepiness and both physiological drowsiness and performance impairment in a controlled laboratory setting. Eighteen healthy young adults (8 women; MAGE =21.44±3.24years) underwent 40hr of extended wakefulness, completing a bihourly Karolinska Sleepiness Scale (KSS) and 10-min Psychomotor Vigilance Task (PVT). Microsleeps and slow eye movements (SEMs) were scored during the PVT. KSS scores increased 3hr prior to performance impairment (p<.001) and 4-6hr prior to physiological sleepiness (p<.001). There were strong within-subject correlations between KSS and PVT lapses (r=0.75, p<.001) and physiological drowsiness (r>0.60, p<.001). Between-subjects product-moment correlations were more modest but showed a significant positive increase across time awake, suggesting that subjective sleepiness and objective outcomes were more tightly correlated after sleep loss. Cross-correlations showed significant positive correlations at 0-lag (p<.034); however, a high proportion of participants showed maximal correlations at positive lags, suggesting KSS was associated with future objective impairment. Within individuals, subjective sleepiness was highly correlated with objective impairment, between-subject correlations were more modest, possibly due to interindividual vulnerability to sleep loss. These results suggest that subjective sleepiness represents an inbuilt early warning system for subsequent drowsiness and performance impairment.

Comparing School-Aged Refraction Measurements Using the 2WIN-S Portable Refractor in Relation to Cycloplegic Retinoscopy: A Cross-Sectional Study.

Purpose To assess the repeatability and agreement of refractive measurements using 2WIN-S photoscreening with the gold-standard cycloplegic retinoscope refraction. Design Single centre, cross-sectional study. Methods Spherical, cylindrical, axis, and spherical equivalent of 194 bilateral eyes of 97 children were assessed using a retinoscope and 2WIN-S. One week later, another operator repeated the 2WIN-S measurements. The primary outcome measures were to assess the repeatability and agreement between spherical equivalent, J0, and J45 readings of 2WIN-S. The repeatability of measurements was assessed by the within-subject standard deviation (2.77 Sw) and intraclass correlation coefficient (ICC). The agreement between devices was assessed using 95% limits of agreement. The extent of the agreement between cycloplegic retinoscopy and noncycloplegic 2WIN-S measurements was assessed using Bland–Altman analysis. Results The mean age ± SD was 10.3 ± 2.46 year (range, 4–14 years). The sphere, cylinder, and spherical equivalent measurements were found to be consistent with both apparatus (r value >0.86). ICC for SE, J0, and J45 was 0.900, 0.666, and 0.639, respectively; Sw for SE, J0, and J45 was 0.61D, 0.30D, and 0.31D, respectively; Bland–Altman analysis of retinoscopy with cycloplegia and 2WIN-S for SE was 184/194 (95%) in 95% confidence interval, and the mean value was 0.46. J0 was 184/194 (95%), and the mean value is −0.04. J45 was 181/194 (93%), and the mean value is −0.15. Conclusion The objective refractive measurement of 2WIN-S had good reliability and high agreement with the gold-standard retinoscopy refraction in children and adolescents. While consistency was observed, it is essential to take into consideration that it is a screening tool.

Improved composite quantile regression and variable selection with nonignorable dropouts

With nonignorable dropouts and outliers, we propose robust statistical inference and variable selection methods for linear quantile regression models based on composite quantile regression and empirical likelihood (EL) that accommodate both the within-subject correlations and nonignorable dropouts. The purpose of our study is threefold. First, we apply the generalized method of moments to estimate the parameters in the nonignorable dropout propensity based on an instrument. Subsequently, the inverse probability weighting and kernel smoothing approaches are applied to obtain the smoothed and bias-corrected generalized estimating equations. Second, we borrow the idea of quadratic inference function to construct the improved EL procedure for nonignorable dropouts. The asymptotic properties of the proposed estimators and their confidence regions are derived. Third, the penalized EL method and algorithm for variable selection are investigated. The finite-sample performance of the proposed estimators is studied through simulation, and an application to HIV-CD4 data set is also presented.