Dataset Information

Comparing baseline and longitudinal measures in association studies.

ABSTRACT: In recent years, longitudinal family-based studies have had success in identifying genetic variants that influence complex traits in genome-wide association studies. In this paper, we suggest that longitudinal analyses may contain valuable information that can enable identification of additional associations compared to baseline analyses. Using Genetic Analysis Workshop 18 data, consisting of whole genome sequence data in a pedigree-based sample, we compared 3 methods for the genetic analysis of longitudinal data to an analysis that used baseline data only. These longitudinal methods were (a) longitudinal mixed-effects model; (b) analysis of the mean trait over time; and (c) a 2-stage analysis, with estimation of a random intercept in the first stage and regression of the random intercept on a single-nucleotide polymorphism at the second stage. All methods accounted for the familial correlation among subjects within a pedigree. The analyses considered common variants with minor allele frequency above 5% on chromosome 3. Analyses were performed without knowledge of the simulation model. The 3 longitudinal methods showed consistent results, which were generally different from those found by using only the baseline observation. The gene CACNA2D3, identified by both longitudinal and baseline approaches, had a stronger signal in the longitudinal analysis (p = 2.65 × 10(-7)) compared to that in the baseline analysis (p = 2.48 × 10(-5)). The effect size of the longitudinal mixed-effects model and mean trait were higher compared to the 2-stage approach. The longitudinal results provided stable results different from that using 1 observation at baseline and generally had lower p values.

SUBMITTER: Wang S

PROVIDER: S-EPMC4143666 | biostudies-literature | 2014

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Comparing baseline and longitudinal measures in association studies.

Wang Shuai S Gao Wei W Ngwa Julius J Allard Catherine C Liu Ching-Ti CT Cupples L Adrienne LA

BMC proceedings 20140617 Suppl 1 Genetic Analysis Workshop 18Vanessa Olmo

In recent years, longitudinal family-based studies have had success in identifying genetic variants that influence complex traits in genome-wide association studies. In this paper, we suggest that longitudinal analyses may contain valuable information that can enable identification of additional associations compared to baseline analyses. Using Genetic Analysis Workshop 18 data, consisting of whole genome sequence data in a pedigree-based sample, we compared 3 methods for the genetic analysis of ...[more]

PMID: 25519412

Similar Datasets

Project description:ImportanceDespite the increasing prevalence of pregravid obesity, systematic evaluation of the association of maternal obesity with fetal growth trajectories is lacking.ObjectiveTo characterize differences in fetal growth trajectories between obese and nonobese pregnant women, and to identify the timing of any observed differences.Design, setting, and participantsThe Eunice Kennedy Shriver National Institute of Child Health and Human Development Fetal Growth Studies-Singletons study enrolled cohorts of pregnant women at 12 US health care institutions. Obese women (with prepregnancy body mass index > 30) and nonobese women (prepregnancy body mass indexes, 19-29.9) without major chronic diseases were recruited between 8 weeks and 0 days' gestation and 13 weeks and 6 days' gestation. A mixed longitudinal randomization scheme randomized participants into 1 of 4 schedules for 2-dimensional and 3-dimensional ultrasonograms to capture weekly fetal growth data throughout the remainder of their pregnancies.Main outcomes and measuresOn each ultrasonogram, fetal humerus length, femur length, biparietal diameter, head circumference, and abdominal circumference were measured. Fetal growth curves were estimated using linear mixed models with cubic splines. Median differences in the fetal measures at each gestational week of the obese and nonobese participants were examined using the likelihood ratio and Wald tests after adjustment for maternal characteristics.ResultsThe study enrolled 468 obese and 2334 nonobese women between 8 weeks and 0 days' gestation and 13 weeks and 6 days' gestation. After a priori exclusion criteria, 443 obese and 2320 nonobese women composed the final cohort. Commencing at 21 weeks' gestation, femur length and humerus length were significantly longer for fetuses of obese woman than those of nonobese women. Differences persisted in obese and nonobese groups through 38 weeks' gestation (median femur length, 71.0 vs 70.2 mm; P = .01; median humerus length, 62.2 vs 61.6 mm; P = .03). Averaged across gestation, head circumference was significantly larger in fetuses of obese women than those of nonobese women (P = .02). Fetal abdominal circumference was not greater in the obese cohort than in the nonobese cohort but was significantly larger than in fetuses of normal-weight women (with body mass indexes between 19.0-24.9) commencing at 32 weeks (median, 282.1 vs 280.2 mm; P = .04). Starting from 30 weeks' gestation, estimated fetal weight was significantly larger for the fetuses of obese women (median, 1512 g [95% CI, 1494-1530 g] vs 1492 g [95% CI, 1484-1499 g]) and the difference grew as gestational age increased. Birth weight was higher by almost 100 g in neonates born to obese women than to nonobese women (mean, 3373.2 vs 3279.5 g).Conclusions and relevanceAs early as 32 weeks' gestation, fetuses of obese women had higher weights than fetuses of nonobese women. The mechanisms and long-term health implications of these findings are not yet established.

Project description:RationaleUnderstanding the reliability and repeatability of clinical measurements used in the diagnosis, treatment and monitoring of disease progression is of critical importance across all disciplines of clinical practice and in clinical trials to assess therapeutic efficacy and safety.ObjectivesOur goal is to understand normal variability for assessing true changes in health status and to more accurately utilize this data to differentiate disease characteristics and outcomes.MethodsOur study is the first study designed entirely to establish the repeatability of a large number of instruments utilized for the clinical assessment of COPD in the same subjects over the same period. We utilized SPIROMICS participants (n = 98) that returned to their clinical center within 6 weeks of their baseline visit to repeat complete baseline assessments. Demographics, spirometry, questionnaires, complete blood cell counts (CBC), medical history, and emphysema status by computerized tomography (CT) imaging were obtained.ResultsPulmonary function tests (PFTs) were highly repeatable (ICC's >0.9) but the 6 minute walk (6MW) was less so (ICC = 0.79). Among questionnaires, the Saint George's Respiratory Questionnaire (SGRQ) was most repeatable. Self-reported clinical features, such as exacerbation history, and features of chronic bronchitis, often produced kappa values <0.6. Reported age at starting smoking and average number of cigarettes smoked were modestly repeatable (kappa = 0.76 and 0.79). Complete blood counts (CBC) variables produced intraclass correlation coefficients (ICC) values between 0.6 and 0.8.ConclusionsPFTs were highly repeatable, while subjective measures and subject recall were more variable. Analyses using features with poor repeatability could lead to misclassification and outcome errors. Hence, care should be taken when interpreting change in clinical features based on measures with low repeatability. Efforts to improve repeatability of key clinical features such as exacerbation history and chronic bronchitis are warranted.

Project description:Longitudinal intervention studies with repeated measurements over time are an important type of experimental design in biomedical research. Due to the advent of “omics”-sciences (genomics, transcriptomics, proteomics, metabolomics), longitudinal studies generate increasingly multivariate outcome data. Analysis of such data must take both the longitudinal intervention structure and multivariate nature of the data into account. The ASCA+-framework combines general linear models with principal component analysis and can be used to separate and visualize the multivariate effect of different experimental factors. However, this methodology has not yet been developed for the more complex designs often found in longitudinal intervention studies, which may be unbalanced, involve randomized interventions, and have substantial missing data. Here we describe a new methodology, repeated measures ASCA+ (RM-ASCA+), and show how it can be used to model metabolic changes over time, and compare metabolic changes between groups, in both randomized and non-randomized intervention studies. Tools for both visualization and model validation are discussed. This approach can facilitate easier interpretation of data from longitudinal clinical trials with multivariate outcomes. Author summary Clinical trials are increasingly generating large amounts of complex biological data. Examples can include measuring metabolism or gene expression in tissue or blood sampled repeatedly over the course of a treatment. In such cases, one might wish to compare changes in not one, but hundreds, or thousands of variables simultaneously. In order to effectively analyze such data, both the study design and the multivariate nature of the data should be considered during data analysis. ANOVA simultaneous component analysis+ (ASCA+) is a statistical method which combines general linear models with principal component analysis, and provides a way to separate and visualize the effects of different factors on complex biological data. In this work, we describe how repeated measures linear mixed models, a class of models commonly used when analyzing changes over time and treatment effects in longitudinal studies, can be used together with ASCA+ for analyzing clinical trials in a novel method called repeated measures-ASCA+ (RM-ASCA+).

Project description:Identification of the majority of organisms present in human-associated microbial communities is feasible with the advent of high throughput sequencing technology. As substantial variability in microbiota communities is seen across subjects, the use of longitudinal study designs is important to better understand variation of the microbiome within individual subjects. Complex study designs with longitudinal sample collection require analytic approaches to account for this additional source of variability. A common approach to assessing community changes is to evaluate the change in alpha diversity (the variety and abundance of organisms in a community) over time. However, there are several commonly used alpha diversity measures and the use of different measures can result in different estimates of magnitude of change and different inferences. It has recently been proposed that diversity profile curves are useful for clarifying these differences, and may provide a more complete picture of the community structure. However, it is unclear how to utilize these curves when interest is in evaluating changes in community structure over time. We propose the use of a bi-exponential function in a longitudinal model that accounts for repeated measures on each subject to compare diversity profiles over time. Furthermore, it is possible that no change in alpha diversity (single community/sample) may be observed despite the presence of a highly divergent community composition. Thus, it is also important to use a beta diversity measure (similarity between multiple communities/samples) that captures changes in community composition. Ecological methods developed to evaluate temporal turnover have currently only been applied to investigate changes of a single community over time. We illustrate the extension of this approach to multiple communities of interest (i.e., subjects) by modeling the beta diversity measure over time. With this approach, a rate of change in community composition is estimated. There is a need for the extension and development of analytic methods for longitudinal microbiota studies. In this paper, we discuss different approaches to model alpha and beta diversity indices in longitudinal microbiota studies and provide both a review of current approaches and a proposal for new methods.

Dataset Information

Comparing baseline and longitudinal measures in association studies.

Publications

Comparing baseline and longitudinal measures in association studies.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets