A Protocol for Characterizing Food Eating and Purchasing Behaviors Using Wearable Sensors to Better Understand Food Acculturation in Hispanics/Latinos (P13-021-19)

Abstract

ObjectivesObservational and cohort studies have examined the eating habits of Hispanics/Latinos often with a focus on types of foods consumed (e.g., sugary beverages, vegetables) to understand how food choices may contribute to obesity, diabetes, and metabolic syndrome in this population. Research has to a lesser extent measured eating behaviors such as eating with family, eating in front of a screen, and other behaviors that could play a significant role in obesity-related outcomes. Furthermore, such behaviors are closely tied to cultural and familial traditions that may morph as Hispanic/Latinos spend more time in the US or over generations. Eating behaviors are notoriously difficult to capture through self-report mechanisms; recent developments in wearable sensors such as cameras or Global Positioning Systems (GPS) can provide objective measurements of what/with who/when/where/how an individual engages with food. This study presents a protocol for collecting, coding, and validating data on food-related behavior with wearable sensors in a population of Hispanics/Latinos. MethodsAn observational cohort of Mexican adults (n = 80) engaged in a one-week study during which they wore a wearable camera (FrontRow) and GPS device. The GPS device logged location (latitude/longitude) every minute. The FrontRow camera recorded images ever 15–30 seconds depending on movement of participant. A protocol was developed to code each image for several food-related behaviors including: social/individual eating, distracted eating, duration/time of eating, snacking, home cooked meals, traditional Mexican meals, and grocery/fast food/convenience/restaurant food purchasing. Validation was performed by second and third coders coding the images and measuring for agreement. Behaviors were linked to locations using time stamp of image and GPS and tied to geographic information system data about food environments to aid in validation of coded images. Resultsn/a. ConclusionsUnderstanding eating and purchasing behaviors could aid in understanding relationships between what Hispanic/Latino individuals eat and obesity, diabetes, and metabolic syndrome outcomes. This protocol provides researchers with a way to objectively measure food-related behaviors that may be connected to food acculturation in a Hispanic/Latino cohort. Funding SourcesNSF, UCSD Qualcomm Institute.