?Efficient statistical methods for assessing dementia risk in Parkinson's disease? Summary/Abstract: The proposed R01 grant is in response to PAR-16-260 ?Methodology and Measurement in the Behavioral and Social Sciences (R01)?. Disease-modifying therapies targeting Parkinson's disease (PD) dementia are likely to be most efficacious before significant cognitive decline has occurred, as has been proposed for Alzheimer's disease (AD). Thus, cognitive biomarker studies in PD are significant because biomarkers may signal an increased risk of future cognitive decline prior to measurable impairment on standard neuropsychological testing. Longitudinal design is particularly desirable because it allows ongoing monitoring of pathophysiological processes associated with cognition and identification of those biomarkers most sensitive to ongoing or future cognitive decline. A major challenge in longitudinal biomarker studies is the difficulty in obtaining all biomarker outcomes serially for every participant, due to limitations in study resources and priorities. Current available statistical procedures such as mixed-effects models ignore missing data, which results in low efficiency (power) of the analyses in the presence of missing data. Thus, our ability to detect significant longitudinal changes in biomarkers is limited by the current available statistical methods due to this inefficiency. This R01 aims to develop more efficient longitudinal methods than the current available methods in the presence of missing biomarker outcome or covariate data. The new methods will require less biomarker data than current methods to achieve the same analytic statistical power (efficiency). This will be a significant methodological advance, as it will reduce future study costs and patient burden without sacrificing power. It has broad applications in PD dementia and other neurodegenerative diseases such as AD, as well as general biomedical research. We also plan to study progression of three potential cognitive biomarkers (cerebrospinal fluid [CSF], brain MRIs, and dopamine transporter [DAT] SPECT imaging) and establish their temporal ordering in relationship to cognitive decline in PD participants in the Parkinson's Progression Markers Initiative (PPMI) study by applying these new statistical methods. The results will inform the design of future studies testing possible disease-modifying therapies in treating PD dementia.

Public Health Relevance

?Efficient statistical methods for assessing dementia risk in Parkinson's disease? Project Narrative: This R01 aims to develop more efficient longitudinal methods than the current available methods in the presence of missing biomarker outcome or covariate data. This will be a significant methodological advance, as it will reduce future study costs and patient burden without sacrificing power. It will improve future biomedical and public health research.

Agency
National Institute of Health (NIH)
Institute
National Institute of Neurological Disorders and Stroke (NINDS)
Type
Research Project (R01)
Project #
5R01NS102324-03
Application #
9691538
Study Section
Special Emphasis Panel (ZRG1)
Program Officer
Babcock, Debra J
Project Start
2017-07-01
Project End
2021-04-30
Budget Start
2019-05-01
Budget End
2020-04-30
Support Year
3
Fiscal Year
2019
Total Cost
Indirect Cost
Name
University of Pennsylvania
Department
Biostatistics & Other Math Sci
Type
Schools of Medicine
DUNS #
042250712
City
Philadelphia
State
PA
Country
United States
Zip Code
19104
Zee, Jarcy; Xie, Sharon X (2018) The Kaplan-Meier Method for Estimating and Comparing Proportions in a Randomized Controlled Trial with Dropouts. Biostat Epidemiol 2:23-33
Rennert, Lior; Xie, Sharon X (2018) Cox regression model with doubly truncated data. Biometrics 74:725-733