Butterfly Girls Study
The study titled ‘Using Technology to Prevent Obesity Among African American Girls’., also known as Butterfly Girls, was a culturally and developmentally appropriate obesity prevention intervention tailored to African American girls. It was presented as an online 8-episode program and followed a three-group randomized design: story and goal setting (treatment), story only (comparison), and waitlist control. The primary outcome was body mass index. Secondary outcomes included child diet and physical activity. The data is currently being examined to determine the influence of the intervention on the outcomes and moderation by socioeconomic status and severe obesity at baseline. This study was conducted via funds from the National Institute on Minority Health and Health Disparities grant #MD005814 (to Dr. Thompson). More information is found on the PI of the study, Dr. Deb Thompson.
Comparison of Body Composition Measures to Assess Childhood Obesity
There are limitations to using body mass index in children to assess unhealthy weight gain and risk for disease such as asthma and cardiovascular disease. We began a systematic literature review to examine the nature of different measures used to assess body composition in children and correlate them with objective measures of body composition.
Markers of Inflammation and Stress in Mothers and Infants
We are conducting a pilot study to measure markers of inflammation and stress in human milk and other biological samples. We intend to correlate the markers with measures of maternal stress, maternal and child diet, and risk for cardiovascular disease.
Social and Demographic Determinants of Maternal and Child Health
We are investigating how maternal perceptions, cultural influences, and attitudes about feeding relate to child growth patterns under the lens of racial/ethnic and sociodemographic disparities. We are also studying how disparities in response to obesity interventions and obesity clinical care within racial/ethnic minority groups can be used to develop strategies for more effective interventions.
We are also conducting secondary data analysis of publicly available social determinants of health data to develop data-driven tools to advance the field of health disparities. The ultimate goal is to promote healthy lifestyle strategies that are effective among low-income mothers with children at-risk for obesity. This will help inform efforts to promote sustainable health-promoting interventions in vulnerable populations.
Our research is grounded in methods development as we frequently encounter analytical challenges during the course of a study. This includes applying methods to deal with imbalanced data in new areas, machine learning methods to answer health disparities questions and advance health equity in obesity research, augmenting quantitative with qualitative approaches, and expanding the use of simulations and multilevel models into fields that traditionally use other methods.
Dr. Musaad was awarded a National Institutes of Health AIM-AHEAD Fellowship Program in Leadership grant for her proposal, "Using Machine Learning to Investigate Intervention Effect Disparities and Digital Phenotypes Among Black Girls; Brain #67674-I." As part of this Fellowship she will apply machine learning methods to the Butterfly Girls data set to extract intra-child and inter-child patterns for activity, sedentary and sleep behaviors. She will integrate the patterns with dietary, anthropometric, demographic, and home environment data to identify disparities in response to the intervention. She will also identify and characterize digital phenotypes that classify children based on their interaction with the online obesity intervention program, in order to relate to behavior change.