Covariance models for nested repeated measures data: analysis of ovarian steroid secretion data

Park, Taesung; Lee, Young Jack

doi:10.1002/sim.949

Cited by 32 publications

(33 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…First, there are multiple segments measured for each subject; second, in some of the patients with asthma, measurements are made at two distinct points in time (before and after bronchial thermoplasty). Linear mixedeffects models for nested repeated measures (38) are used to examine differences in SVP between patients with asthma and healthy volunteers, as well as within patients with asthma in whom measurements were made both before and after bronchial thermoplasty. This Tables E1 and E2 (online), and a bar plot of segmental data is shown in Figure 4.…”

Section: Methodsmentioning

confidence: 99%

Regional Ventilation Changes in Severe Asthma after Bronchial Thermoplasty with³He MR Imaging and CT

Thomen¹,

Sheshadri²,

Quirk³

et al. 2015

Radiology

122

View full text Add to dashboard Cite

).q RSNA, 2014 Purpose:To quantify regional lung ventilation in healthy volunteers and patients with severe asthma (both before and after thermoplasty) by using a combination of helium 3 ( 3 He) magnetic resonance (MR) imaging and computed tomography (CT), with the intention of developing more effective image-guided treatments for obstructive lung diseases. Materials andMethods:With approval of the local institutional review board, informed consent, and an Investigational New Drug Exemption, six healthy volunteers and 10 patients with severe asthma were imaged in compliance with HIPAA regulations by using both multidetector CT and 3He MR imaging. Individual bronchopulmonary segments were labeled voxel by voxel from the CT images and then registered to the Results:Ventilation measures for healthy volunteers yielded smaller segment-to-segment variation (mean SVP, 100% 6 18 [standard deviation]) than did the measures for patients with severe asthma (mean SVP, 97% 6 23). Patients with asthma also demonstrated larger segmental defect percentages (median, 13.5%; interquartile range, 8.9%-17.8%) than healthy volunteers (median, 6%; interquartile range, 5.6%-6.3%). These quantitative results confirm what is visually observed on the 3 He images. A Spearman correlation of r = 20.82 was found between the change in whole-lung defect percentage and the number of days between final treatment and second 3 He imaging. Conclusion:Regional quantification of lung ventilation is indeed feasible and may be a useful technique for image-guided treatment of obstructive lung diseases, such as bronchial thermoplasty for severe asthma. In these patients, ventilation defects decreased as a function of time after treatment.q RSNA, 2014

show abstract

Section: Methodsmentioning

confidence: 99%

Regional Ventilation Changes in Severe Asthma after Bronchial Thermoplasty with³He MR Imaging and CT

Thomen¹,

Sheshadri²,

Quirk³

et al. 2015

Radiology

122

View full text Add to dashboard Cite

show abstract

“…With multilevel models, a covariance matrix that reflects dependence between observations can be chosen. Thus, modeling the within-subjects covariance structure is particularly relevant, since the accuracy of the regression parameter estimates depends on the right choice (Littell, Pendergast, & Natarajan, 2000;Park & Lee, 2002).…”

Section: Integrating the Two Levels In The Lmmmentioning

confidence: 99%

Using the linear mixed model to analyze nonnormal data distributions in longitudinal designs

et al. 2012

View full text Add to dashboard Cite

Using a Monte Carlo simulation and the KenwardRoger (KR) correction for degrees of freedom, in this article we analyzed the application of the linear mixed model (LMM) to a mixed repeated measures design. The LMM was first used to select the covariance structure with three types of data distribution: normal, exponential, and log-normal. This showed that, with homogeneous between-groups covariance and when the distribution was normal, the covariance structure with the best fit was the unstructured population matrix. However, with heterogeneous between-groups covariance and when the pairing between covariance matrices and group sizes was null, the best fit was shown by the between-subjects heterogeneous unstructured population matrix, which was the case for all of the distributions analyzed. By contrast, with positive or negative pairings, the within-subjects and betweensubjects heterogeneous first-order autoregressive structure produced the best fit. In the second stage of the study, the robustness of the LMM was tested. This showed that the KR method provided adequate control of Type I error rates for the time effect with normally distributed data. However, as skewness increased-as occurs, for example, in the log-normal distribution-the robustness of KR was null, especially when the assumption of sphericity was violated. As regards the influence of kurtosis, the analysis showed that the degree of robustness increased in line with the amount of kurtosis. Keywords Longitudinal data . Linear mixed model . Kenward-Roger method . Robustness . Nonnormal distributionsOver the last three decades, the analysis of repeated measures data has centered on the linear mixed model (LMM). Laird and Ware (1982) established the basis of the LMM with the incorporation of the within-subjects error correlation. Their work was subsequently extended by Cnaan, Laird, and Slasor (1997) and Verbeke and Molenberghs (2000), who applied the LMM to longitudinal data. In contrast to analyses that are based on variances (ANOVA and MANOVA), the LMM models the structure of the covariance matrix. This enables a more efficient estimation of the fixed effects and, consequently, yields more robust statistical tests. However, when the covariance structures are not properly fitted and the sample sizes are small, the Type I error rate tends to rise (Wright & Wolfinger, 1996).With respect to the covariance structure, Keselman, Algina, Kowalchuk, and Wolfinger (1999) demonstrated that when the covariance matrix is not spherical, the degrees of freedom associated with the conventional F test are too large. One way of controlling this bias is to apply the procedure developed by Kenward and Roger (1997; henceforth, the KR method) to correct the degrees of freedom. Studies by Kowalchuk, Keselman, Algina, and Wolfinger (2004) and Vallejo and Ato (2006) showed that with an adequate covariance structure, the KR method is able, in

show abstract

“…[13] introduced the use of the Kronecker product to construct covariance matrices for bi-dimensional spatial data. Such covariance structures have also been used for data that have two repeated factors, such as days nested within menstrual cycles [15]. Kronecker product covariance matrices are also useful for multivariate data that have more than one level of complexity (e.g., multivariate outcomes, with each outcome measured repeatedly) since each level can be modelled separately.…”

Section: Modelling the Covariancementioning

confidence: 99%

Regression models for the analysis of longitudinal Gaussian data from multiple sources

O’Brien

Fitzmaurice

2005

Statistics in Medicine

View full text Add to dashboard Cite

We present a regression model for the joint analysis of longitudinal multiple source Gaussian data. Longitudinal multiple source data arise when repeated measurements are taken from two or more sources, and each source provides a measure of the same underlying variable and on the same scale. This type of data generally produces a relatively large number of observations per subject; thus estimation of an unstructured covariance matrix often may not be possible. We consider two methods by which parsimonious models for the covariance can be obtained for longitudinal multiple source data. The methods are illustrated with an example of multiple informant data arising from a longitudinal interventional trial in psychiatry.

show abstract

Covariance models for nested repeated measures data: analysis of ovarian steroid secretion data

Cited by 32 publications

References 10 publications

Regional Ventilation Changes in Severe Asthma after Bronchial Thermoplasty with³He MR Imaging and CT

Regional Ventilation Changes in Severe Asthma after Bronchial Thermoplasty with³He MR Imaging and CT

Using the linear mixed model to analyze nonnormal data distributions in longitudinal designs

Regression models for the analysis of longitudinal Gaussian data from multiple sources

Contact Info

Product

Resources

About

Covariance models for nested repeated measures data: analysis of ovarian steroid secretion data

Cited by 32 publications

References 10 publications

Regional Ventilation Changes in Severe Asthma after Bronchial Thermoplasty with3He MR Imaging and CT

Regional Ventilation Changes in Severe Asthma after Bronchial Thermoplasty with3He MR Imaging and CT

Using the linear mixed model to analyze nonnormal data distributions in longitudinal designs

Regression models for the analysis of longitudinal Gaussian data from multiple sources

Contact Info

Product

Resources

About

Regional Ventilation Changes in Severe Asthma after Bronchial Thermoplasty with³He MR Imaging and CT

Regional Ventilation Changes in Severe Asthma after Bronchial Thermoplasty with³He MR Imaging and CT