The CoLab score is associated with SARS-CoV-2 viral load during admission in individuals admitted to the intensive care unit: the CoLaIC cohort study.

OBJECTIVES: The present study examines the temporal association between the changes in SARS-CoV-2 viral load during infection and whether the CoLab-score can facilitate de-isolation. METHODS: Nasal swabs and blood samples were collected from ICU-admitted SARS-CoV-2 positive patients at Maastricht UMC+ from March 25, 2020 to October 1, 2021. The CoLab-score was calculated based on 10 blood parameters and age and can range from -43 to 6. Three mixed effects analyses compared patient categories based on initial PCR Ct values (low; Ct≤20, mid; 20>Ct≤30, high; Ct>30), serial PCR Ct values to CoLab-scores over time, and the association between within-patient delta Ct values and CoLab-scores. RESULTS: In 324 patients, the median Ct was 33, and the median CoLab-score was -1.78. Mid (n=110) and low (n=41) Ct-categories had higher CoLab-scores over time (+0.60 points, 95 % CI; 0.04-1.17, and +0.28 points, 95 % CI -0.49 to 1.04) compared to the high Ct (n=87) category. Over time, higher serial Ct values were associated with lower serial CoLab-scores, decreasing by -0.07 points (95 % CI; -0.11 to -0.02) per day. Increasing delta Ct values were associated with a decreasing delta CoLab-score of -0.12 (95 % CI; -0.23; -0.01). CONCLUSIONS: The study found an association between lower viral load on admission and reduced CoLab-score. Additionally, a decrease in viral load over time was associated with a decrease in CoLab-score. Therefore, the CoLab-score may make patient de-isolation an option based on the CoLab-score.

Minimal detectable change of gait and balance measures in older neurological patients: estimating the standard error of the measurement from before-after rehabilitation data thanks to the linear mixed-effects models.

BACKGROUND: Tracking gait and balance impairment in time is paramount in the care of older neurological patients. The Minimal Detectable Change (MDC), built upon the Standard Error of the Measurement (SEM), is the smallest modification of a measure exceeding the measurement error. Here, a novel method based on linear mixed-effects models (LMMs) is applied to estimate the standard error of the measurement from data collected before and after rehabilitation and calculate the MDC of gait and balance measures. METHODS: One hundred nine older adults with a gait impairment due to neurological disease (66 stroke patients) completed two assessment sessions before and after inpatient rehabilitation. In each session, two trials of the 10-meter walking test and the Timed Up and Go (TUG) test, instrumented with inertial sensors, have been collected. The 95% MDC was calculated for the gait speed, TUG test duration (TTD) and other measures from the TUG test, including the angular velocity peak (ωpeak) in the TUG test's turning phase. Random intercepts and slopes LMMs with sessions as fixed effects were used to estimate SEM. LMMs assumptions (residuals normality and homoscedasticity) were checked, and the predictor variable ln-transformed if needed. RESULTS: The MDC of gait speed was 0.13 m/s. The TTD MDC, ln-transformed and then expressed as a percentage of the baseline value to meet LMMs' assumptions, was 15%, i.e. TTD should be < 85% of the baseline value to conclude the patient's improvement. ωpeak MDC, also ln-transformed and expressed as the baseline percentage change, was 25%. CONCLUSIONS: LMMs allowed calculating the MDC of gait and balance measures even if the test-retest steady-state assumption did not hold. The MDC of gait speed, TTD and ωpeak from the TUG test with an inertial sensor have been provided. These indices allow monitoring of the gait and balance impairment, which is central for patients with an increased falling risk, such as neurological old persons. TRIAL REGISTRATION: NA.

Enrollment forecast for clinical trials at the portfolio planning phase based on site-level historical data.

An accurate forecast of a clinical trial enrollment timeline at the planning phase is of great importance to both corporate strategic planning and trial operational excellence. The naive approach often calculates an average enrollment rate from historical data and generates an inaccurate prediction based on a linear trend with the average rate. Under the traditional framework of a Poisson-Gamma model, site activation delays are often modeled with either fixed initiation time or a simple random distribution while incorporating the user-provided site planning information to achieve good forecast accuracy. However, such user-provided information is not available at the early portfolio planning stage. We present a novel statistical approach based on generalized linear mixed-effects models and the use of non-homogeneous Poisson processes through the Bayesian framework to model the country initiation, site activation, and subject enrollment sequentially in a systematic fashion. We validate the performance of our proposed enrollment modeling framework based on a set of 25 preselected studies from four therapeutic areas. Our modeling framework shows a substantial improvement in prediction accuracy in comparison to the traditional statistical approach. Furthermore, we show that our modeling and simulation approach calibrates the data variability appropriately and gives correct coverage rates for prediction intervals of various nominal levels. Finally, we demonstrate the use of our approach to generate the predicted enrollment curves through time with confidence bands overlaid.

Oxytocin Modifies the Excitability and the Action Potential Shape of the Hippocampal CA1 GABAergic Interneurons.

Oxytocin (OT) is a neuropeptide that modulates social-related behavior and cognition in the central nervous system of mammals. In the CA1 area of the hippocampus, the indirect effects of the OT on the pyramidal neurons and their role in information processing have been elucidated. However, limited data are available concerning the direct modulation exerted by OT on the CA1 interneurons (INs) expressing the oxytocin receptor (OTR). Here, we demonstrated that TGOT (Thr4,Gly7-oxytocin), a selective OTR agonist, affects not only the membrane potential and the firing frequency but also the neuronal excitability and the shape of the action potentials (APs) of these INs in mice. Furthermore, we constructed linear mixed-effects models (LMMs) to unravel the dependencies between the AP parameters and the firing frequency, also considering how TGOT can interact with them to strengthen or weaken these influences. Our analyses indicate that OT regulates the functionality of the CA1 GABAergic INs through different and independent mechanisms. Specifically, the increase in neuronal firing rate can be attributed to the depolarizing effect on the membrane potential and the related enhancement in cellular excitability by the peptide. In contrast, the significant changes in the AP shape are directly linked to oxytocinergic modulation. Importantly, these alterations in AP shape are not associated with the TGOT-induced increase in neuronal firing rate, being themselves critical for signal processing.

Individual and Family Determinants for Quality of Life in Parents of Children with Inborn Errors of Metabolism Requiring a Restricted Diet: A Multilevel Analysis Approach.

OBJECTIVE: The objective of this study was to compare the quality of life (QoL) for parents of children with inborn errors of metabolism (IEMs) requiring a restricted diet with French population norms and investigate parental QoL determinants. STUDY DESIGN: This cross-sectional study included mothers and/or fathers of children < 18 years of age affected by IEMs requiring a restricted diet (except phenylketonuria) from January 2015 to December 2017. Parents' QoL was assessed using the World Health Organization Quality of Life BREF questionnaire and compared with age- and sex-matched reference values from the French general population. Linear mixed models were used to examine the effects of demographic, socioeconomic, disease-related, and psychocognitive factors on parental QoL, according to a 2-level regression model considering individuals (parents) nested within families. RESULTS: Of the 1156 parents invited to participate, 785 (68%) were included. Compared with the general population, parents of children with IEMs requiring a restricted diet reported a lower QoL in physical and social relationship domains but a higher QoL in the psychological domain. In the multivariate analysis, characteristics associated with poorer parental QoL included both parent-related factors (being a father, older age, more educated parent, nonworking parent, greater anxiety, seeking more social support, and using less positive thinking and problem-solving coping strategies) and family-related factors (disease complications, increased number of hospital medical providers, child's younger age, single-parent family, and lower family material wealth). CONCLUSION: Parents of children with IEMs requiring a restricted diet reported poorer QoL in physical and social relationship domains than population norms. Psychocognitive factors, beyond disease-specific and family-related characteristics, were the most important determinants influencing parental QoL and may represent essential aspects for interventions. CLINICAL TRIAL REGISTRATION: ClinicalTrials.gov: NCT02552784.

Model-based clustering of high-dimensional longitudinal data via regularization.

We propose a model-based clustering method for high-dimensional longitudinal data via regularization in this paper. This study was motivated by the Trial of Activity in Adolescent Girls (TAAG), which aimed to examine multilevel factors related to the change of physical activity by following up a cohort of 783 girls over 10 years from adolescence to early adulthood. Our goal is to identify the intrinsic grouping of subjects with similar patterns of physical activity trajectories and the most relevant predictors within each group. The previous analyses conducted clustering and variable selection in two steps, while our new method can perform the tasks simultaneously. Within each cluster, a linear mixed-effects model (LMM) is fitted with a doubly penalized likelihood to induce sparsity for parameter estimation and effect selection. The large-sample joint properties are established, allowing the dimensions of both fixed and random effects to increase at an exponential rate of the sample size, with a general class of penalty functions. Assuming subjects are drawn from a Gaussian mixture distribution, model effects and cluster labels are estimated via a coordinate descent algorithm nested inside the Expectation-Maximization (EM) algorithm. Bayesian Information Criterion (BIC) is used to determine the optimal number of clusters and the values of tuning parameters. Our numerical studies show that the new method has satisfactory performance and is able to accommodate complex data with multilevel and/or longitudinal effects.

Analysis of therapy monitoring in the International Congenital Adrenal Hyperplasia Registry.

OBJECTIVE: Congenital adrenal hyperplasia (CAH) requires exogenous steroid replacement. Treatment is commonly monitored by measuring 17-OH progesterone (17OHP) and androstenedione (D4). DESIGN: Retrospective cohort study using real-world data to evaluate 17OHP and D4 in relation to hydrocortisone (HC) dose in CAH patients treated in 14 countries. PATIENTS: Pseudonymized data from children with 21-hydroxylase deficiency (21OHD) recorded in the International CAH Registry. MEASUREMENTS: Assessments between January 2000 and October 2020 in patients prescribed HC were reviewed to summarise biomarkers 17OHP and D4 and HC dose. Longitudinal assessment of measures was carried out using linear mixed-effects models (LMEM). RESULTS: Cohort of 345 patients, 52.2% female, median age 4.3 years (interquartile range: 3.1-9.2) were taking a median 11.3 mg/m2 /day (8.6-14.4) of HC. Median 17OHP was 35.7 nmol/l (3.0-104.0). Median D4 under 12 years was 0 nmol/L (0-2.0) and above 12 years was 10.5 nmol/L (3.9-21.0). There were significant differences in biomarker values between centres (p < 0.05). Correlation between D4 and 17OHP was good in multiple regression with age (p < 0.001, R2 = 0.29). In longitudinal assessment, 17OHP levels did not change with age, whereas D4 levels increased with age (p < 0.001, R2 = 0.08). Neither biomarker varied directly with dose or weight (p > 0.05). Multivariate LMEM showed HC dose decreasing by 1.0 mg/m2 /day for every 1 point increase in weight standard deviation score. DISCUSSION: Registry data show large variability in 17OHP and D4 between centres. 17OHP correlates with D4 well when accounting for age. Prescribed HC dose per body surface area decreased with weight gain.

Online two-way estimation and inference via linear mixed-effects models.

In this article, we tackle the estimation and inference problem of analyzing distributed streaming data that is collected continuously over multiple data sites. We propose an online two-way approach via linear mixed-effects models. We explicitly model the site-specific effects as random-effect terms, and tackle both between-site heterogeneity and within-site correlation. We develop an online updating procedure that does not need to re-access the previous data and can efficiently update the parameter estimate, when either new data sites, or new streams of sample observations of the existing data sites, become available. We derive the non-asymptotic error bound for our proposed online estimator, and show that it is asymptotically equivalent to the offline counterpart based on all the raw data. We compare with some key alternative solutions both analytically and numerically, and demonstrate the advantages of our proposal. We further illustrate our method with two data applications.

Impact of model misspecification on model-based tests in PK studies with parallel design: real case and simulation studies.

This article evaluates the performance of pharmacokinetic (PK) equivalence testing between two formulations of a drug through the Two-One Sided Tests (TOST) by a model-based approach (MB-TOST), as an alternative to the classical non-compartmental approach (NCA-TOST), for a sparse design with a few time points per subject. We focused on the impact of model misspecification and the relevance of model selection for the reference data. We first analysed PK data from phase I studies of gantenerumab, a monoclonal antibody for the treatment of Alzheimer's disease. Using the original rich sample data, we compared MB-TOST to NCA-TOST for validation. Then, the analysis was repeated on a sparse subset of the original data with MB-TOST. This analysis inspired a simulation study with rich and sparse designs. With rich designs, we compared NCA-TOST and MB-TOST in terms of type I error and study power. With both designs, we explored the impact of misspecifying the model on the performance of MB-TOST and adding a model selection step. Using the observed data, the results of both approaches were in general concordance. MB-TOST results were robust with sparse designs when the underlying PK structural model was correctly specified. Using the simulated data with a rich design, the type I error of NCA-TOST was close to the nominal level. When using the simulated model, the type I error of MB-TOST was controlled on rich and sparse designs, but using a misspecified model led to inflated type I errors. Adding a model selection step on the reference data reduced the inflation. MB-TOST appears as a robust alternative to NCA-TOST, provided that the PK model is correctly specified and the test drug has the same PK structural model as the reference drug.

Blood Analytes as Biomarkers of Mechanisms Involved in Alzheimer's Disease Progression.

Alzheimer's disease (AD) is the leading cause of dementia, but the pathogenetic factors are not yet well known, and the relationships between brain and systemic biochemical derangements and disease onset and progression are unclear. We aim to focus on blood biomarkers for an accurate prognosis of the disease. We used a dataset characterized by longitudinal findings collected over the past 10 years from 90 AD patients. The dataset included 277 observations (both clinical and biochemical ones, encompassing blood analytes encompassing routine profiles for different organs, together with immunoinflammatory and oxidative markers). Subjects were grouped into four severity classes according to the Clinical Dementia Rating (CDR) Scale: mild (CDR = 0.5 and CDR = 1), moderate (CDR = 2), severe (CDR = 3) and very severe (CDR = 4 and CDR = 5). Statistical models were used for the identification of potential blood markers of AD progression. Moreover, we employed the Pathfinder tool of the Reactome database to investigate the biological pathways in which the analytes of interest could be involved. Statistical results reveal an inverse significant relation between four analytes (high-density cholesterol, total cholesterol, iron and ferritin) with AD severity. In addition, the Reactome database suggests that such analytes could be involved in pathways that are altered in AD progression. Indeed, the identified blood markers include molecules that reflect the heterogeneous pathogenetic mechanisms of AD. The combination of such blood analytes might be an early indicator of AD progression and constitute useful therapeutic targets.

Outcome-Stratified Analysis of Biomarker Trajectories for Patients Infected With Severe Acute Respiratory Syndrome Coronavirus 2.

Longitudinal trajectories of vital signs and biomarkers during hospital admission of patients with COVID-19 remain poorly characterized despite their potential to provide critical insights about disease progression. We studied 1884 patients with severe acute respiratory syndrome coronavirus 2 infection from April 3, 2020, to June 25, 2020, within 1 Maryland hospital system and used a retrospective longitudinal framework with linear mixed-effects models to investigate relevant biomarker trajectories leading up to 3 critical outcomes: mechanical ventilation, discharge, and death. Trajectories of 4 vital signs (respiratory rate, ratio of oxygen saturation (Spo2) to fraction of inspired oxygen (Fio2), pulse, and temperature) and 4 laboratory values (C-reactive protein (CRP), absolute lymphocyte count (ALC), estimated glomerular filtration rate, and D-dimer) clearly distinguished the trajectories of patients with COVID-19. Before any ventilation, log(CRP), log(ALC), respiratory rate, and Spo2-to-Fio2 ratio trajectories diverge approximately 8-10 days before discharge or death. After ventilation, log(CRP), log(ALC), respiratory rate, Spo2-to-Fio2 ratio, and estimated glomerular filtration rate trajectories again diverge 10-20 days before death or discharge. Trajectories improved until discharge and remained unchanged or worsened until death. Our approach characterizes the distribution of biomarker trajectories leading up to competing outcomes of discharge versus death. Moving forward, this model can contribute to quantifying the joint probability of biomarkers and outcomes when provided clinical data up to a given moment.

Evaluative judgment across domains: Liking balance, contour, symmetry and complexity in melodies and visual designs.

Evaluative judgment-i.e., assessing to what degree a stimulus is liked or disliked-is a fundamental aspect of cognition, facilitating comparison and choosing among alternatives, deciding, and prioritizing actions. Neuroimaging studies have shown that evaluative judgment involves the projection of sensory information to the reward circuit. To investigate whether evaluative judgments are based on modality-specific or modality-general attributes, we compared the extent to which balance, contour, symmetry, and complexity affect liking responses in the auditory and visual modalities. We found no significant correlation for any of the four attributes across sensory modalities, except for contour. This suggests that evaluative judgments primarily rely on modality-specific sensory representations elaborated in the brain's sensory cortices and relayed to the reward circuit, rather than abstract modality-general representations. The individual traits art experience, openness to experience, and desire for aesthetics were associated with the extent to which design or compositional attributes influenced liking, but inconsistently across sensory modalities and attributes, also suggesting modality-specific influences.

Not water, sanitation and hygiene practice, but timing of stunting is associated with recovery from stunting at 24 months: results from a multi-country birth cohort study.

OBJECTIVES: To measure the role of water, sanitation and hygiene (WASH) practices on recovery from stunting and assess the role of timing of stunting on the reversal of this phenomenon. DESIGN: Data from the MAL-ED multi-country birth cohort study was used for the current analysis. Generalised linear mixed-effects models were used to estimate the probability of reversal of stunting with WASH practice and timing of stunting as the exposures of interest. SETTING: Seven different countries across three continents. PARTICIPANTS: A total of 612 children <2 years of age. RESULTS: We found that not WASH practice but timing of stunting had statistically significant association with recovery from stunting. In comparison with the children who were stunted at 6 months, children who were stunted at 12 months had 1·9 times (ß = 0·63, P = 0·03) more chance of recovery at 24 months of age. And, children who were stunted at 18 months of age even had higher odds (adjusted OR = 3·01, ß = 1·10, P < 0·001) of recovery than children who were stunted at 6 months. Additionally, mother's height (ß = 0·59, P = 0·04) and household income (ß = 0·02, P < 0·05) showed statistically significant associations with the outcome. CONCLUSIONS: The study provided evidence for the role of timing of stunting on the recovery from the phenomenon. This novel finding indicates that the programmes to promote linear growth should be directed at the earliest possible timepoints in the course of life.

Investigation of the relationship between heart rate and functional class in pulmonary hypertension.

The utility of a non-linear mixed effects model to model heart rate is examined. The heart rate acceleration is derived as a parameter from this model. The relationship between different potential measures of disease severity including heart rate acceleration is examined. Our study is focused on heart rate variability (HRV), heart rate acceleration (HRA), oxygen saturation (SaO2) and the six-minute walk distance (6MWD) as well as their relationship to WHO functional class. The results and conclusions are derived from data were collected by the Children Hospital of Colorado.

Clinical factors affecting pupillary light reflex parameters: a single-centre, cross-sectional study.

PURPOSE: To evaluate the effects of stimulus intensity, aging, sex, smoking and eye symmetry on pupillary light reflex (PLR) parameters. METHODS: We evaluated 2812 eyes from 1406 subjects in a single-centre, cross-sectional study. PLR data were collected using four different stimulus intensities. We prepared two models for each of the eight PLR parameters, and defined the model with the lowest values of Akaike's information criterion (AIC) as being the best-fit. Model A was a linear regression model without adjustment for among-individual variability, while the Model B linear mixed-effects models (LMMs) were adjusted for among-individual variability. The regression coefficients of the two models were compared. RESULTS: Model B showed the lowest AIC values for all parameters and the best fit. For light stimulus intensity, age and eye symmetry, the two models yielded similar results for all PLR parameters. For sex and smoking index, some PLR parameters showed the opposite results, i.e., Model A showed significant effects while Model B did not. CONCLUSION: These results indicate that light stimulus intensity, aging, sex, smoking and eye symmetry are factors that affect PLR parameters. These should be adjusted when evaluating the clinical potential of PLR as a diagnostic tool. In addition, adjusting for among-individual variability due to LMMs can improve the model fit and reduce false positives. This can reveal the association between clinical factors and PLR parameters with increased accuracy.

Low agreement among reviewers evaluating the same NIH grant applications.

Obtaining grant funding from the National Institutes of Health (NIH) is increasingly competitive, as funding success rates have declined over the past decade. To allocate relatively scarce funds, scientific peer reviewers must differentiate the very best applications from comparatively weaker ones. Despite the importance of this determination, little research has explored how reviewers assign ratings to the applications they review and whether there is consistency in the reviewers' evaluation of the same application. Replicating all aspects of the NIH peer-review process, we examined 43 individual reviewers' ratings and written critiques of the same group of 25 NIH grant applications. Results showed no agreement among reviewers regarding the quality of the applications in either their qualitative or quantitative evaluations. Although all reviewers received the same instructions on how to rate applications and format their written critiques, we also found no agreement in how reviewers "translated" a given number of strengths and weaknesses into a numeric rating. It appeared that the outcome of the grant review depended more on the reviewer to whom the grant was assigned than the research proposed in the grant. This research replicates the NIH peer-review process to examine in detail the qualitative and quantitative judgments of different reviewers examining the same application, and our results have broad relevance for scientific grant peer review.

Exact variance component tests for longitudinal microbiome studies.

In metagenomic studies, testing the association between microbiome composition and clinical outcomes translates to testing the nullity of variance components. Motivated by a lung human immunodeficiency virus (HIV) microbiome project, we study longitudinal microbiome data by using variance component models with more than two variance components. Current testing strategies only apply to models with exactly two variance components and when sample sizes are large. Therefore, they are not applicable to longitudinal microbiome studies. In this paper, we propose exact tests (score test, likelihood ratio test, and restricted likelihood ratio test) to (a) test the association of the overall microbiome composition in a longitudinal design and (b) detect the association of one specific microbiome cluster while adjusting for the effects from related clusters. Our approach combines the exact tests for null hypothesis with a single variance component with a strategy of reducing multiple variance components to a single one. Simulation studies demonstrate that our method has a correct type I error rate and superior power compared to existing methods at small sample sizes and weak signals. Finally, we apply our method to a longitudinal pulmonary microbiome study of HIV-infected patients and reveal two interesting genera Prevotella and Veillonella associated with forced vital capacity. Our findings shed light on the impact of the lung microbiome on HIV complexities. The method is implemented in the open-source, high-performance computing language Julia and is freely available at https://github.com/JingZhai63/VCmicrobiome.

Collision between biological process and statistical analysis revealed by mean centring.

Animal ecologists often collect hierarchically structured data and analyse these with linear mixed-effects models. Specific complications arise when the effect sizes of covariates vary on multiple levels (e.g. within vs. among subjects). Mean centring of covariates within subjects offers a useful approach in such situations, but is not without problems. A statistical model represents a hypothesis about the underlying biological process. Mean centring within clusters assumes that the lower level responses (e.g. within subjects) depend on the deviation from the subject mean (relative) rather than on the absolute scale of the covariate. This may or may not be biologically realistic. We show that mismatch between the nature of the generating (i.e. biological) process and the form of the statistical analysis produce major conceptual and operational challenges for empiricists. We explored the consequences of mismatches by simulating data with three response-generating processes differing in the source of correlation between a covariate and the response. These data were then analysed by three different analysis equations. We asked how robustly different analysis equations estimate key parameters of interest and under which circumstances biases arise. Mismatches between generating and analytical equations created several intractable problems for estimating key parameters. The most widely misestimated parameter was the among-subject variance in response. We found that no single analysis equation was robust in estimating all parameters generated by all equations. Importantly, even when response-generating and analysis equations matched mathematically, bias in some parameters arose when sampling across the range of the covariate was limited. Our results have general implications for how we collect and analyse data. They also remind us more generally that conclusions from statistical analysis of data are conditional on a hypothesis, sometimes implicit, for the process(es) that generated the attributes we measure. We discuss strategies for real data analysis in face of uncertainty about the underlying biological process.

Evaluation of non-linear-mixed-effect modeling to reduce the sample sizes of pediatric trials in type 2 diabetes mellitus.

Recruitment for pediatric trials in Type II Diabetes Mellitus (T2DM) is very challenging, necessitating the exploration of new approaches for reducing the sample sizes of pediatric trials. This work aimed at assessing if a longitudinal Non-Linear-Mixed-Effect (NLME) analysis of T2DM trial could be more powerful and thus require fewer patients than two standard statistical analyses commonly used as primary or sensitivity efficacy analysis: Last-Observation-Carried-Forward (LOCF) followed by (co)variance (AN(C)OVA) analysis at the evaluation time-point, and Mixed-effects Model Repeated Measures (MMRM) analysis. Standard T2DM efficacy studies were simulated, with glycated hemoglobin (HbA1c) as the main endpoint, 24 weeks' study duration, 2 arms, assuming a placebo and a treatment effect, exploring three different scenarios for the evolution of HbA1c, and accounting for a dropout phenomenon. 1000 trials were simulated, then analyzed using the 3 analyses, whose powers were compared. As expected, the longitudinal modeling MMRM analysis was found to be more powerful than the LOCF + ANOVA analysis at week 24. The NLME analysis gave slightly more accurate drug-effect estimations than the two other methods, however it tended to slightly overestimate the magnitude of the drug effect, and it was more powerful than the MMRM analysis only in some scenarios of slow HbA1c decrease. The gain in power afforded by NLME was more apparent when two additional assessments enriched the design; however, the gain was not systematic for all scenarios. Finally, this work showed that NLME analyses may help to reduce significantly the required sample sizes in T2DM pediatric studies, but only for enriched designs and slow HbA1c decrease.

Stochastic Lanczos estimation of genomic variance components for linear mixed-effects models.

BACKGROUND: Linear mixed-effects models (LMM) are a leading method in conducting genome-wide association studies (GWAS) but require residual maximum likelihood (REML) estimation of variance components, which is computationally demanding. Previous work has reduced the computational burden of variance component estimation by replacing direct matrix operations with iterative and stochastic methods and by employing loose tolerances to limit the number of iterations in the REML optimization procedure. Here, we introduce two novel algorithms, stochastic Lanczos derivative-free REML (SLDF_REML) and Lanczos first-order Monte Carlo REML (L_FOMC_REML), that exploit problem structure via the principle of Krylov subspace shift-invariance to speed computation beyond existing methods. Both novel algorithms only require a single round of computation involving iterative matrix operations, after which their respective objectives can be repeatedly evaluated using vector operations. Further, in contrast to existing stochastic methods, SLDF_REML can exploit precomputed genomic relatedness matrices (GRMs), when available, to further speed computation. RESULTS: Results of numerical experiments are congruent with theory and demonstrate that interpreted-language implementations of both algorithms match or exceed existing compiled-language software packages in speed, accuracy, and flexibility. CONCLUSIONS: Both the SLDF_REML and L_FOMC_REML algorithms outperform existing methods for REML estimation of variance components for LMM and are suitable for incorporation into existing GWAS LMM software implementations.