Thanksgiving offers data frenzy for obesity researchers

It’s Thanksgiving. The turkey is carved, potatoes are adequately mashed, and a buffet of pumpkin, pecan and apple pies lines the counter ready to be sliced as soon as the dinner plates have been cleared.

You think to yourself, “I don’t want to fill up on stuffing — so I’ll just have a little and save room for pumpkin pie. With a scoop of ice cream. And maybe a sliver of apple pie, as well. I mean, it’s Thanksgiving.”

You have a plan.

But as soon as you sit down, Aunt Nancy asks a seemingly benign question about your relationship status, and your mom chimes in that you’re still single and paying off college loans. Suddenly the reasonable scoop of potatoes on your plate has transformed into a mountain, and half of the marshmallows that top the candied yam casserole have somehow made it onto your plate. Your hand — as if possessed — keeps shoveling the tasteless food into your mouth.

What happened?

According to Donna Spruijt-Metz, professor of psychology (research) and director of the Mobile Health (mHealth) Collaboratory at USC’s Center for Social and Economic Research (CESR), most obesity researchers throw out their data from food-based holidays like Thanksgiving and Christmas, deeming it anomalous. But for Spruijt-Metz and her research team, these holidays generate invaluable information.

“Thanksgiving — even if it’s a really joyful situation — is a little fraught for most people,” she said. “As a behavioral health researcher and someone interested in family interactions around food, I see the holiday as creating a data frenzy.”

Spruijt-Metz and Kayla De La Haye, assistant professor of preventive medicine at Keck School of Medicine of USC, have teamed up with researchers at the University of Virginia’s School of Engineering and Applied Science to devise a data-driven method for combating obesity. But instead of focusing on what people eat, the study will observe family dynamics while eating — and what triggers overeating. 

Donna Spruijt-Metz, professor of psychology (research).

“How we interact around food — eating too fast, eating with someone who is overweight, being around someone who irritates you — all of that impacts eating behavior, and in turn obesity,” Spruijt-Metz said. “And we also know we can intervene on those factors.”

The new study, deemed M2FED (an acronym for “monitoring and modeling the family eating dynamics”) has received a $1.7 million grant from the National Science Foundation (NSF). Researchers will provide wearable sensors to 40 volunteer families and place sensors in their homes to watch how their emotions and interactions around food affect eating behavior.

University of Virginia researchers John Lach and John Stankovic will devise a cyber-physical system (CPS) of sensors to monitor factors such as tone of voice, distractions during mealtimes, frequency of eating and meals and stress levels of participants.

Cyber-physical systems refer to a network of interacting devices — in this case participants’ smart phones, smart watches and a number of sensors placed strategically around their homes.

“We will also be able to ask them questions such as whether they are hungry when they are eating and monitor their feelings in real time,” Spruijt-Metz said. “Imagine coming home from a stressful day. Your first instinct might be to grab a handful of chocolates from the cabinet, even if dinner is in an hour.

“But what if we, as psychologists and engineers, could find ways to use new mobile technologies to help people stop themselves from doing that in real-time and in the context in which they are tempted and need help?”

Once they have clear data models to show how interactions influence eating, Spruijt-Metz and De La Haye will begin testing different strategies people can employ to avoid overeating or eating mindlessly.

“There is not a lot known about the effectiveness of specific momentary interventions,” Spruijt-Metz said. “With this technology, we can watch these interactions in real time, and give what we call a ‘just-in-time adaptive intervention.’ It could be as simple as flashing a good cat video on your phone after a stressful meeting to keep you from grabbing and eating all the treats in the break room.”

Although the project is in its nascent stages of development — they plan to deploy the equipment in the first volunteer home within the next three months — Spruijt-Metz already has high hopes.

“I believe that if we can ask the right questions at the right time, we may be able to give people a little nudge to keep them from engaging in behaviors that affect their eating for life,” Spruijt-Metz said. “In the long run, I think this approach is much more holistic and sustainable than trying to manage obesity by trying to measure and intervene on exactly what people eat.”