|
Activity Number:
|
374
|
|
Type:
|
Invited
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Social Statistics Section
|
| Abstract - #300262 |
|
Title:
|
Improved Analysis of Weight-Loss Interventions for African-American Women
|
|
Author(s):
|
Justine Shults*+ and Xiaoying Wu and Shiriki Kumanyika
|
|
Companies:
|
University of Pennsylvania and University of Pennsylvania and University of Pennsylvania
|
|
Address:
|
610 Blockley Hall, School of Medicine, Philadelphia, PA, 19104-6021,
|
|
Keywords:
|
familial correlation structure ; longitudinal study ; quasi-least squares ; health disparities ; generalized estimating equations ; correlated data
|
|
Abstract:
|
Longitudinal studies can address health disparities in diet related health conditions affecting minority populations. We consider a study that assesses involvement of family or friends as a strategy for enhancing social support and success in weight loss in African Americans. The need to analyze family/friend clusters provides motivation for application of a familial correlation structure that previously was not available for generalized estimating equations (GEE). In general, by expanding the list of structures for GEE, we aim to improve efficiency, which could enable improved detection of clinically important differences by ethnic subgroup, or effects when the drop out rate is higher than expected, or the sample size is limited initially. This research was completed as part of an NIH funded program to address characteristics typical of longitudinal studies in minority populations.
|
