Longitudinal Analysis for Diverse Populations Project

The “Longitudinal Analysis for Diverse Populations” (LADP) project R01CA096885 was generously funded by a grant from the National Cancer Institute (NCI) in the National Institutes of Health. This completed project was based at the Center of Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, and was led by Justine Shults.

Jump to:     Abstract     Publications     Software     Presentations     Investigators

Abstract

(Also available on the NCI web-site.)

Community based interventions (e.g. to reduce obesity and increase physical activity) can play an important role in reducing the risk and overall mortality and morbidity of diseases such as coronary heart disease and cancer. They are especially important for African Americans, who are disproportionately at risk for a wide range of negative health conditions, including cancer of the breast, colon, esophagus, prostate, pancreas, ant stomach; mortality from cardiovascular disease; hypertension; and elevated serum cholesterol. This project will develop more efficient and cost-effective methods for analysis of longitudinal studies using quasi-least squares (QLS), with special emphasis on studies in diverse populations. Our aims are: (1) To develop more efficient and informative methods for analysis in longitudinal studies and community-based interventions, by applying QLS for a wide range of correlation models not currently implemented for generalized estimating equations (GEE) and constructing confidence intervals and tests of hypotheses for the parameters of the new structures, for data with one or more levels of within-cluster associations (e.g. both within families and within subjects over time). (2) To develop methods for planning more powerful studies and taking advantage of re-computing interim power, by (i) assessing loss in efficiency in estimation for different study designs and correlation models and (ii) providing explicit formulas for sample size and power calculations for several correlation structures, including the structures implemented in Aim 1. This aim will consider both the regression and the correlation parameters. (3) To apply our methods in analyses of several studies in female, pediatric, and African-American Populations at the University of Pennsylvania, to further refine and tailor their development to the characteristics of data for diverse populations and to answer new questions that our methods make possible. (4) To compare and contrast our approaches with alternative methods, including methods based on random effects models and recent extensions of GEE, via simulations to assess small sample efficiency and bias and data analyses to compare results of the different approaches. (5) To implement the methods for analysis (Aim 1) and planning (Aim 2) in Stata programs, for use by other statisticians. Further, to widely disseminate the programs, and their documentation, on a web site developed for this project.

mso-border-top-alt:solid windowtext .5pt;padding:0.5pt 0in 0in 0in;margin-left: .25in;margin-right:21.0pt'> Jump to:     Abstract     Publications     Software     Presentations     Investigators

Publications

1. Justine Shults, Carissa A. Mazurick, J. Richard Landis. Analysis of repeated bouts of measurements in the framework of generalized estimating equations. Statistics in Medicine, 25(23): 4114-4128, 2006.

2. Justine Shults, Wenguang Sun, Xin Tu, and Jay Amsterdam. (2006) On the Violation of bounds for the correlation in Generalized Estimating Equation analyses Of binary data from longitudinal trials. UPenn Biostatistics Working Papers, Working Paper 8. [Available from Bepress]

3. Wenguang Sun, Justine Shults, and Mary Leonard. A Note on the use of unbiased estimating equations to estimate correlation in analysis of longitudinal trials. Biometrical Journal, 51(1): 5-18, 2009. An earlier version of this paper is also available: (2006) Use of Unbiased Estimating Equations to Estimate Correlation in Generalized Estimating Equation Analysis of Longitudinal Trials, UPenn Biostatistics Working Papers, Working Paper 4. [Available from Bepress]

4. Sarah J. Ratcliffe and Justine Shults. GEEQBOX: A Matlab toolbox for generalized estimating equations and quasi-least squares. The Journal of Statistical Software, 25(14): 1-14, 2008.

5. Justine Shults, Sarah J. Ratcliffe, and Mary Leonard. Improved generalized estimating equation analysis via xtqls for implementation of quasi-least squares in Stata. The Stata Journal, 7(2): 147-166, 2007. [Available from Bepress: UPenn Biostatistics Working Papers, Working Paper 13.]

6. Xin Tu, Jiameng Zhang, Jeanne Kowalski, Justine Shults, Changyong Feng, Wenguang Sun and Wan Tang. (2006) Power analysis for longitudinal study designs with missing data. Statistics in Medicine, 26(15): 2958-81, 2007. [Note: Justine Shults was supported by the LADP project. The other authors were supported by the following grants and contracts: R01-DA012249, AI-51186, and NIH contract N01-AI-50029.]

7. Justine Shults and Sarah J. Ratcliffe. Analysis of multi-level correlated data in the framework of generalized estimating equations via xtmultcorr procedures in Stata and qls functions in Matlab. Statistics and its Interface, 2: 187-196, 2009. An earlier version of this paper is also available: (2007) UPenn Biostatistics Working Papers, Working Paper 15. [Available from Bepress]

8. Jichun Xie and Justine Shults. Implementation of quasi-least squares with the R package qlspack. Journal of Statistical Software. (In press)

9. Hanjoo Kim, Joseph Hilbe, and Justine Shults. (2008). On the designation of the patterned associations for longitudinal Bernoulli data: Weight matrix versus true correlation structure? (under review at Biometrika) UPenn Biostatistics Working Papers, Working Paper 26. [Available from Bepress]

10. Hanjoo Kim and Justine Shults. %QLS SAS Macro: A SAS macro for Analysis of Longitudinal Data Using Quasi-Least Squares. Journal of Statistical Software, 35(2), 2010. An earlier version of this paper is also available: (2008) UPenn Biostatistics Working Papers, Working Paper 27. [Available from Bepress]

mso-border-top-alt:solid windowtext .5pt;padding:0.5pt 0in 0in 0in;margin-left: .25in;margin-right:21.0pt'> Jump to:     Abstract     Publications     Software     Presentations     Investigators

Software

Software for this project is described and made available in the following publications. By using this software, you agree to cite the appropriate publication in any manuscripts. To install the software, simply unzip the files into the working directory for the analysis. Changing to that directory within Stata or Matlab will make the procedures / functions available.

GEEQBOX: A Matlab toolbox for implementation of quasi-least squares as described in Publication #4 (Ratcliffe & Shults, 2008). [Download software and User's Guide in a .zip file]

xtqls: Stata procedure for implementation of quasi-least squares as described in Publication #5 (Shults, Ratcliffe & Leonard, 2007). [Download software in a .zip file]

xtmultcorr: Stata procedures for the analysis of multi-level correlated data as described in Publication #7 (Shults & Ratcliffe, 2007). [Download software in a .zip file]

qlspack: R package for the analysis of correlated data via quasi-least squares as described in Publication #8 (Xie & Shults, 2008). [Available on CRAN web-site]

%QLS: A SAS macro for analysis of longitudinal data via quasi-lease squares as described in Publication #10 (Kim & Shults, 2008). [Download software in a .zip file]

Please write to jshults at mail.med.upenn.edu (Stata), jichun at mail.med.upenn.edu (R software), or sratclif at upenn.edu (Matlab software) to request software for the familial correlation structure

mso-border-top-alt:solid windowtext .5pt;padding:0.5pt 0in 0in 0in;margin-left: .25in;margin-right:21.0pt'> Jump to:     Abstract     Publications     Software     Presentations     Investigators

Presentations

Invited Presentations (Date, Title and Location, Speaker)

September 2004,   “Analysis of Repeated Bouts of Measurements in the Framework of Generalized Estimating Equations”, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio. Justine Shults

November 2004,   “Analysis of Repeated Bouts of Measurements in the Framework of Generalized Estimating Equations”, Department of Mathematics and Statistics, Georgia State University, Atlanta, GA. Justine Shults

January 2005,   “Analysis of Dichotomous Outcomes Using Quasi-Least Squares”, Department of Mathematics and Statistics, Oregon State University, Corvallis, OR. Justine Shults

March 2006,   “A Generalized Estimating Equation Analysis of a Clinical Trial to Compare Venlafaxine with Lithium in the Treatment of Major Depressive Episode”, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio. Justine Shults

September 2006,   “Improved Generalized Estimating Equation Analysis via Application of Quasi-Least Squares”, Department of Biostatistics, Virginia Commonwealth University, Richmond, Virginia. Justine Shults

November 2006,   “Improved Generalized Estimating Equation Analysis via Application of Quasi-Least Squares”, Department of Statistics, George Washington University, Washington D.C. Justine Shults

March 2007,   “On the Impact and Likelihood of a Violation of Bounds for the Correlation in GEE Analyses of Binary Data for Longitudinal Data and What we Can Do to Address this Problem", ENAR, Atlanta, GA. Justine Shults, Wenguang Sun, Xin Tu, and Tom Ten Have.

August 2008,   “Improved Analysis of Weight-Loss Interventions for African-American Women”, Joint Statistical Meetings (JSM), Denver, Colorado. Justine Shults, Xiaoying Wu, and Shriiki Kumanyika.

Contributed Presentations (Date, Title and Location, Speaker)

March 2005,   “Adjusted Quasi-Least Squares for Valid Analysis of Correlated Binary Data”, ENAR, Austin, TX; Justine Shults and Wenguang Sun.

March 2005,   “Comparison of Wang-Carey Estimation Versus Quasi-Least Squares”, ENAR, Austin, TX; Wenguang Sun and Justine Shults.

August 2005,   “Adjusted Quasi-Least Squares for Analysis of Correlated Binary Data”, JSM, Minneapolis, MN; Justine Shults and Wenguang Sun.

August 2005,   “Comparison of Wang-Carey Estimation Versus Quasi-Least Squares”, JSM, Minneapolis, MN; Wenguang Sun and Justine Shults.

March 2007,   “Implementation of a New Correlation Structure in Framework of GEE with R Software", ENAR, Atlanta, GA; Jichun Xie and Justine Shults.

August 2007,   “Implementation of an Extended Familial Correlation Structure with Quasi-Least Squares", JSM, Salt Lake City, UT; Jichun Xie and Justine Shults.

August 2008,   “Demonstration of User-Written Software in Stata, R, and Matlab for Analysis of a Familial Correlation Structure in a Quasi-Least Squares Analysis of Weight Loss in the SHARE Study", JSM, Denver, CO; Xiaoying Wu, Jichun Xie, Sarah Ratcliffe, Shiriki Kumanyika, and Justine Shults.

mso-border-top-alt:solid windowtext .5pt;padding:0.5pt 0in 0in 0in;margin-left: .25in;margin-right:21.0pt'> Jump to:     Abstract     Publications     Software     Presentations     Investigators

Investigators

Investigators on this project include the following faculty members in the Department of Biostatistics and Epidemiology, Center for Clinical Epidemiology and Biostatistics (CCEB), University of Pennsylvania's Perelman School of Medicine, Philadelphia, PA.

Investigator Role on Project
Justine Shults, PhD Principal Investigator
Scarlett Bellamy, ScD Co-Investigator
Shiriki Kumanyika, PhD, MPH Co-Investigator
Sarah J. Ratcliffe, PhD Co-Investigator
Thomas Ten Have, PhD, MPH Co-Investigator
Jinbo Chen, PhD Co-Investigator (joined in year 3)

Graduate Students in the Department of Biostatistics within the CCEB who have worked on this project include:

  • Wenguang Sun, MS completed his Master’s Project in Biostatistics on problems stemming from this project. Publication #3 describes his research; plus see the list of contributed talks for two presentations of this research, at ENAR 2005 and JSM 2005. (Update: Wenguang Sun, PhD is currently an Assistant Professor on the tenure track in the Department of Statistics at North Carolina State University.)
  • Jichun Xie, BS studied at Peking University in Bejing, China. She joined the LADP project during year 3 of the project.
  • Hanjoo Kim, BS studied at George Washington University. He joined the LADP project during year 4 of the project.

Student Interns who have worked on this project include:

  • Marina Y. Shishova (Summer 2007), a Mathematics and Statistics major at Mount Holyoke College.
  • George Wang (Summer 2008), a Statistics major at Penn State University
mso-border-top-alt:solid windowtext .5pt;padding:1.0pt 0in 0in 0in;margin-left: .25in;margin-right:21.0pt'>