08 Nov Brain Gray Matter Volume Predictive of Weight Loss Success
MedicalResearch.com Interview with:
Fatemeh Mokhtari
Medical Imaging PhD Student
VT-WFU SBES
MedicalResearch.com: What is the background for this study? What are the main findings?
Response: The objective of this study was to use baseline anatomical brain MRI scans to prospectively predict weight loss success following an intensive lifestyle intervention. In the study, 52 participants, age 60 to 79, were recruited from the Cooperative Lifestyle Interventions Programs II (CLIP-II) project. The participants were overweight or obese (BMI greater than 28 and less than 42) and had a history of either cardiovascular disease or metabolic syndrome. All participants had a baseline MRI scan, and then were randomized to one of three groups – diet only, diet plus aerobic exercise training or diet plus resistance exercise training. The goal of the 18-month diet and exercise program was a weight loss of 7 to 10 percent of body mass.
Basic brain structure information garnered from the MRIs was classified using a support vector machine, a type of computerized predictive algorithm. Specifically, we trained a computational predictive model which mapped each subject’s brain scan to weight loss performance. Predictions were based on baseline brain gray and white matter volume from the participants’ MRIs and compared to the study participants’ actual weight loss after the 18 months. The accuracy of the model was then tested, and our prediction algorithms were 78% accurate in predicting successful weight loss. Brain gray matter volume provided higher prediction accuracy compared with white matter and the combination of the two outperformed either one alone.
MedicalResearch.com: What should readers take away from your report?
Response: We have developed a simple test that can predict weight loss success using an anatomical brain MRI scan, and we hope that this test could ultimately be used to tailor treatment for patients—real personalized medicine. For example, people identified at high risk for failure might benefit from intensive treatment and close guidance. People identified as having a high probability for success might require less intensive treatment.
MedicalResearch.com: What recommendations do you have for future research as a result of this study?
Response: The current study lays the foundation of our future studies which will include larger sample size and separate analyses for males and females, since gender is a determinant factor in brain function and self-regulating behaviors. Our preliminary study with small sample size prevented us from independently predicting weight loss in males and females. We are also now adding brain functional images and various health and cognitive measures to the anatomical images in our predictive model and believe our future models will be even more accurate.
The ultimate goal of our research is to create several predictive models, each specific to a weight loss treatment, such as exercise, diet, and pharmaceutical interventions. Following a brain MRI scan, the patient would receive the treatment for which the highest weight loss performance is predicted. We plan to develop a user-friendly software package, and clinicians would be able to specify treatment protocols for each individual. Such personalized treatments will play a key part in the practice of medicine in the future.
MedicalResearch.com: Is there anything else you would like to add?
Response: Every day you see advertisements pop up on your computer or smart phone screens, which are related to your previous online shopping. Social networks and online shopping companies predict your future behaviors from your previous clicks. Such predictions are similar to what we do in this study. We aim to predict how overweight patients will perform in lifestyle-based treatments for weight loss. This is achieved by looking at their brain as a major determinant of behavior. Our prediction model looks for the optimal relationship between brain anatomy and weight loss performance and then is applied to future subjects. Our model could assist clinicians to design the most efficient obesity treatment for each patient; this would be best for the patients and would reduce healthcare costs.
Co-authors are: Fatemeh Mokhtari, M.Sc., Brielle M. Paolini, Ph.D., Jonathan H. Burdette, M.D., and Paul Laurienti, M.D., Ph.D., of Wake Forest Baptist; and W. Jack Rejeski, Ph.D., and Anthony P. Marsh, Ph.D., of Wake Forest University.
MedicalResearch.com: Thank you for your contribution to the MedicalResearch.com community.
Citation:
Note: Content is Not intended as medical advice. Please consult your health care provider regarding your specific medical condition and questions.
More Medical Research Interviews on MedicalResearch.com
[wysija_form id=”5″]
Last Updated on November 8, 2016 by Marie Benz MD FAAD