| Project Type: | Pilot Study | ||
| Project Title: | Using baseline brain imaging to predict success in weight loss interventions in older adults | ||
| Principal Investigator: | Paul Laurienti, MD, PhD | ||
| Other Key Investigator (1): | |||
| Other Key Investigator (2): | |||
| Other Key Investigator (3): | |||
| Other Key Investigator (4): | |||
| Project Start Date: | 04/01/2016 | ||
| Projected End Date: | 03/31/2018 | ||
| Associated Center: | Wake Forest University | ||
| Associated Core(s): |
Pilot and Exploratory Studies Core |
||
| Brief Description: | We propose to use machine-learning applied to baseline brain imaging data to discriminate success with weight loss. The main outcome of interest is the amount of weight lost during the first 6-month intensive phase of treatment with the goal of being able to discriminate between participants who fall into the upper and lower half of this distribution. An exploratory analysis will determine if machine learning can also predict physical function. We will use data from two existing Pepper Center projects that examined weight loss in older adults. The first study (CLIP-II) evaluated mobility disability following a community-based intervention that included weight loss, weight loss + aerobic exercise, and weight loss + resistance training. Brain imaging data was collected on a subset of participants before the intervention and at 6 months. We have begun examining baseline brain imaging data and have promising preliminary analyses using a support vector machine to predict weight loss based on brain anatomy. We anticipate that when combined with functional brain networks, the performance of the classification algorithm will improve substantially. The second study (INFINITE) examined aerobic fitness in older adults following exercise combined with moderate or intensive caloric restriction. Brain imaging data collected on a subset of participants by Dr. Christina Hugenschmidt will be used to determine if the machine learning algorithm developed on the CLIP-II data can be cross-validated in an independent data set. The preliminary data generated by this proposal will position us to submit an R01designed to predict weight loss success, a priori, in older adults participating in weight loss programs and to eventually develop pharmacologic and/or mindfulness-based behavioral interventions to treat this at-risk group. Of relevance to this project is the fact that in December of 2015, NIDDDK conducted a workshop to stimulate interest in investigating phenotypes for weight loss. Dr. Rejeski, a co-investigator on this project, was an invited speaker at that workshop. It is anticipated that an RFA on this topic will be released within the coming year. | ||
| Project Keywords: |
Intervention Images Obesity/Weight Loss |
||
| Registered By: | Abby Archer | ||
| Registered On: | 08/15/2016 | ||
| Project Status: | Active | ||
OAIC Registry of Works-in-Progress
- Home /
- Works-in-Progress /
- Project Details