Gillen Furthers Healthcare Research with Two NIH Grants

Daniel L. Gillen, professor and chair of the Department of Statistics, recently collaborated on two Multiple-Principal-Investigator National Institutes of Health grants, both related to exploiting data analysis to improve healthcare.

The first grant, “Neuroimaging Biomarkers for Cognitive Decline in Elderly with Amyloid Pathology,” is with Dr. Michael Yassa, an associate professor and Chancellor’s Fellow from UCI’s Department of Neurobiology and Behavior. The $3.6 million grant aims to develop and assess neuroimaging biomarkers for predicting cognitive decline. “One of the major initiatives in Alzheimer’s disease is to develop earlier signals for when someone is going to start to decline. … If nothing else, it helps families to prepare,” says Gillen. As co-PI, his job will be to develop novel statistical methods for predicting cognitive decline and apply the work to subject data obtained as part of the grant.

For the other grant, “The Study Partner Requirement in Preclinical Alzheimer’s Disease Trials,” Gillen serves as co-PI with Dr. Joshua Grill, an associate professor of psychiatry and human behavior in the UCI School of Medicine. The two-year grant, with $222,000 awarded for the first year, will focus on the Food and Drug Administration’s study-partner requirement for patients undergoing neuro-cognitive tests, which states that a partner of the patient (such as a spouse, child, or caretaker) must also supply responses to a cognitive test. According to Gillen, “little is known about whether the study partner requirement helps.” Using historical data from two large and well-known studies in Alzheimer’s disease, Grill and Gillen aim to provide empirical evidence to either justify or refute the study-partner requirement by assessing how much additional information and accuracy is gained from the responses of study partners. For this assessment, Gillen will build predictive models based on the patient assessment, the study partner assessment, and both assessments combined to identify which model is most effective in recognizing early cognitive decline.