Association of Obesity and Weight Status with Dementia and Cognition among Older Adults in the U.S.
Faculty Mentor: Dr. Virginia Chang | Graduate Student Mentor: Yuyu (Ruby) Chen
The concurrent and ongoing epidemics of obesity and dementia have elicited serious concern over the role of obesity in dementia. As even a modest effect of obesity on increasing the risk of dementia could have far-reaching public health implications, an understanding of the relationship between weight status and cognitive outcomes is vital to public health. While obesity is a well-known risk factor for cardiovascular disease, diabetes, cancer, and other health outcomes, its relationship to cognition is less well understood. In this project, students will examine the association of weight status and obesity with dementia and cognitive status. We will examine these relationships in cross-section and also in an event history framework with incident cases. We will also consider whether these relationships vary by central sociodemographic factors such as sex, race/ethnicity, education, and income. Analyses will be conducted using the Health and Retirement Study (HRS), which is an ongoing, nationally representative longitudinal survey of the U.S. population over the age of 50. The interviews asked about a wide variety of factors including demographic characteristics, health, and cognitive performance. Underrepresented racial groups are over-sampled, providing a diverse sample of older adults. The HRS is designed and conducted by the University of Michigan, and it is supported by the National Institute on Aging (NIA/NIH) and the Social Security Administration.
NYU GPH Postdoctoral fellow Eunyoung Choi also assisted in this project.
PQAR Students: Abena Dinizulu, Taliyah Griffin, Jessica Sanchez
A Matched Observational Study
Faculty Mentor: Dr. Siyu Heng | Graduate Student Mentor: Jiawei Zhang
Many previous studies found that advanced maternal age at birth is associated with adverse birth outcomes such as low birth weight, child stunting, and preterm birth, especially among populations with limited access to health care. Rigorous causal inference studies are in substantial demand to clarify if these detected associations are due to actual causal effects or confounding bias (i.e., the bias introduced by confounders that are associated with both the treatment and the outcomes of interest). In this project, we will conduct a matched observational study to evaluate the causal impacts of advanced maternal age on birth weight, using the datasets from the Demographic and Health Surveys (DHS) conducted in sub-Saharan African countries. To adjust for confounding bias, we will use optimal matching to match each treated subject (i.e., a mother whose maternal age is above or equal to 35) with one or several controls (i.e., mothers whose maternal age lies between 20 to 29) on measured confounders. We will then conduct nonparametric randomization inference to robustly test the causal impacts of advanced maternal age on birth weight in the matched dataset. We will also discuss the public health implications of our causal inference results and future work.
PQAR Students: Qiyun (Krystal) Liu, Stephanie Perez, Michael Terrefortes Rosado
Faculty Mentor: Dr. Jonathan Odumegwu | Graduate Student Mentor: Kexin Fu
Observational studies found substantial risk reductions of coronary heart disease (CHD) events in postmenopausal women on oral estrogen therapy. However, these studies might suffer from confounding by the potential that women with a more favorable CVD risk profile take estrogen therapy as compared to non-users. To control for that potential confounding, a randomized controlled trial, HERS (Heart and Estrogen/progestin Replacement Study), was conducted; this was designed to test the safety and efficacy of hormone-replacement therapy (estrogen plus progestin, HRT) for the prevention of recurrent coronary heart disease (CHD) events. Women were randomly allocated to either HRT or placebo and then monitored for CHD events.
The purpose of this project is to use the HERS dataset to understand and show an ability to analyze data from a biomedical research study. During the program, we will play and analyze the data from the HERS study, such as univariate and bivariate analyses, fitting appropriate models, and interpreting and communicating the findings.
For more details about the study, see this paper: Grady D, Applegate W, Bush T, Furberg C, Riggs B, Hulley SB. Heart and Estrogen/progestin Replacement Study (HERS): design, methods, and baseline characteristics.
PQAR Students: Adesua Aigbomian, Torriana Avery, Alexa DeVilmè
Tobacco use is a well-established preventable cause of disease, disability, and death in the United States. However, there is a research gap in understanding the patterns of tobacco use among older adults, as most studies have primarily focused on the youth and young adult population. In this project, we aim to address this gap by using longitudinal data from a nationally representative sample of older adults in the United States in the Population Assessment of Tobacco and Health (PATH) study (Waves 1-5 data collected from 2013 to 2018). We aim to gain insights into the temporal nature of tobacco use development among older adults. Specifically, we will analyze data to assess the developmental trajectory of tobacco use (e.g., e-cigarette) prevalence and frequency of use. We will explore the risk factors (e.g., mental health symptoms) associated with tobacco use trajectory. Additionally, if time permits, we may also investigate different subgroups based on their growth trajectory patterns. To accomplish these objectives, we will use the statistical programming language R for all our analyses. Through the project, students will learn longitudinal statistical methods and gain hand-on experience in data management, data exploration, growth curve analysis, as well as growth mixture analysis. By the end of the study, students will have developed proficiency in conducting these analyses and effectively communicating the findings through oral presentations and PowerPoint slides.