Generating novel predictive models to estimate the risk of future ASCVD & Dementia in older adults
Funded Grant
Overview
Affiliation
View All
Overview
description
Response to Grants for Early Medical/Surgical Specialists' Transition to Aging Research (GEMSSTAR) Competition Title: Generating novel predictive models to estimate the risk of future ASCVD & Dementia in older adults Project Summary/Abstract: A person’s baseline risk determines to a large extent their anticipated benefit from many preventive treatments. Older patients desire to live longer while maintaining cognitive function and freedom from dementia, including Alzheimer’s disease, the #1 cause of morbidity and disability in older adults. Older adults also prioritize avoiding atherosclerotic cardiovascular disease (ASCVD), the #1 killer of older adults. Importantly, many risk factors for Alzheimer’s disease and dementia also increase risk for ASCVD. Alzheimer’s disease is the most common etiologic basis for incident mild cognitive impairment and dementia in older adults and can be identified as the cause in 70-75% of cases. Thus, providing older patients with personalized risk estimates for both dementia, including Alzheimer’s disease, and ASCVD could facilitate a comprehensive, evidence-based and patient-centered approach to therapeutic decision making in older adults. Unfortunately, current risk models were derived in younger adults, and fail to accurately predict risk in older adults. Second it remains unclear whether existing ASCVD risk models can also predict dementia risk and vice versa. Finally, to date, no one has evaluated whether these risk estimates help stratify therapeutic benefits of intervention in older adults. Leveraging a mentorship team of world experts in geriatrics, cardiology, and epidemiology, I will utilize data from subjects ≥75 years old from the National Heart, Lung, and Blood Institute (NHLBI) Pooled Cohorts in order to develop a clinical risk model to estimate risk of dementia, including Alzheimer’s disease, at 5 years from the selected baseline visit (Aim 1). In parallel, we will develop a clinical risk model to estimate the risk of ASCVD over the same time period in the same population of individuals ≥75 years old. In addition to traditional risk factors, we will derive these models using a novel set of candidate predictors not previously included in prior risk models including baseline cognition, functional status, depression, and mobility. Both models will then be externally validated using data from the REasons for Geographic And Racial Differences in Stroke (REGARDS) cohort (Aim 2). Finally, we will apply our model to patients ≥75 years old from the Systolic Blood Pressure Intervention Trial (SPRINT) in order to determine whether therapeutic benefit from intensive vs. conservative anti-hypertensive therapy in older adults differs across levels of predicted risk (Aim 3). Once developed and validated, we will develop an electronic health record-based version of the model for widespread dissemination and use in clinical care. The training I will receive through this work will give me expertise in model building and deployment and broaden my research interest in dementia including Alzheimer’s disease. It will also lay the groundwork for a future application for the Paul B. Beeson Emerging Leaders Career Development Award and other independent funding, with the ultimate goal of becoming an independent clinician-researcher focused on the care of older adults.
