Preschoolers in Head Start (HS) come from low-income families, often have working single parents, and are vulnerable to obesity. Approximately 33% of HS children in Houston, Texas are overweight or obese;this is 1.5x higher than the national rate. Energy Intake (EI) is an important predictor of weight gain and accurate methods to measure children's EI are greatly needed. Limitations to accurate food measurements in preschoolers include multiple caregivers, eating outside the home, respondent burden, need for parental recalls, difficulty estimating portion sizes, reporting bias, and cost. Nutritional epidemiology and obesity prevention studies with young children are limited because little progress has been made toward the development of valid and reliable dietary assessment methods that are feasible with an ethnically diverse population of young children in the natural environment. Moving the diet assessment field forward will allow a better understanding of how dietary factors are related to childhood obesity and allow effective evaluation of obesity prevention programs. To overcome the gaps in dietary assessment methodology of preschool children and the unique challenges of this age group, valid, reliable, and age-appropriate measures are needed. A Remote Food Photography Method (RFPM) was developed and validated with adults. For the proposed research, adults (i.e. primary caregivers and HS assistants) will use Smartphones to capture images of preschool children's food selection and plate waste while the children are in their natural environment. These images are sent immediately to the researchers using a wireless network. Consistent with Ecological Momentary Assessment (EMA) methodology, automated prompts are sent to participants'Smartphone's reminding them to capture and send images of the child's foods. A computer program (The Food Photography Application [FPA]) manages the prompts and food images from participants. Images are analyzed to estimate food intake using existing methodology by the FPA, which automatically identifies the foods and estimates food intake using computer imaging algorithms. Participant burden is minimized and the method does not rely on the participant to estimate portion size accurately, which is one of the largest sources of error for self-report methods. The goal of the proposed research is to modify the RFPM for use with preschoolers. Primary caregivers and HS assistants will use a Smartphone to capture 24-hour images of HS children food selections and plate waste. The RFPM, EMA, and FPA will be used to estimate the energy and nutrient intake of preschoolers. The accuracy of the RFPM will be examined by determining the energy and nutrient intake and comparing them to established methods: 1) gram weights of laboratory-based test meals and 2) doubly labeled water in a natural environment. The research is a collaboration between the Children's Nutrition Research Center (CNRC) (PI: T. Nicklas) and the Pennington Biomedical Research Center (PBRC) (Co-I: C. Martin).
Our study has the ability to further improve methodology for assessing dietary and food intake by young children in both laboratory and natural environments. The use of a valid and reliable measure to assess young children's dietary and food intakes has the long term potential use in epidemiologic studies and obesity prevention programs. Our preliminary studies and those of others, suggests that this dietary assessment method will be successful in determining valid dietary intakes in young children.