Widespread use of mobile tech is game-changer

Dr. Bonnie Spring

The ubiquitous nature of the mobile phone is a game-changer for researchers focused on mobile health, Dr. Bonnie Spring of Northwestern University told a group gathered to mark the 20th anniversary of the National Institutes of Health’s Office of Behavioral and Social Sciences Research (OBSSR).

Speaking to a day-long research symposium held in late June, Spring, a professor of preventive medicine, psychology, psychiatry, and public health defined mobile health as health promotion of all kinds supported by mobile devices. It’s the sensors in the mobile devices that have great potential as detectors of changes in a person’s state of risk.

As head of the Northwestern Center for Behavior and Health, Spring’s research focuses on using technology-supported intervention to change behavior and promote healthy lifestyles. She is also an investigator for the Center of Excellence for Mobile Sensor Data-to-Knowledge (MD2K), where she is the lead of the smoking cessation research thrust. MD2K is funded by the National Institutes of Health through its Big Data to Knowledge (BD2K) Initiative.

Behavioral health practitioners already use well-validated change techniques, such as self-monitoring and social support to help patients make healthy changes.  But the addition of mobile technology alters things by expanding the reach of the practitioner, Spring said.

In the U.S., the population averages more than one mobile phone subscription per person. Even in developing countries, like some in Africa where only 25 percent of the population has electricity and only 29 percent of the roads are paved, 60 percent of the population has a mobile phone.

“As someone trained to do behavioral interventions in the same format as a therapy, face-to-face in an office for a 50-minute hour, this is a real game-changer,” Spring said. “Mobile technologies really helps improve access. I don’t have to have the person drive a long distance, pay for parking, arrange child care. Remote communities don’t have to worry about having lack of expert interventionists in their local area.

“If somebody is carrying a mobile device, they carry the intervention tools they need with them. Mobile devices are portable, they can access the tools they need on demand when they’re needed, and all the channels are present,” she said. “A smartphone isn’t just a phone. It’s the internet, your whole social network. It’s a map. It’s a coach. It’s all-in-one, and you’re carrying it with you.”

Spring said that while smoking remains the most significant cause of preventable premature mortality, obesity is a more prevalent problem that also needs to be addressed.

Recent research has used smart phone applications, or apps, to facilitate behavior change related to obesity and smoking.

The risk behaviors associated with obesity include poor diet and physical inactivity. Apps have been developed to help people make healthier food choices and to remind them to be more active.

Related article
Using Sensors to Sort One Gesture from Another

But, when discussing models for interventions to change behavior, “mostly we have a tremendous amount of faith in fostering adherence to self-monitoring of the behaviors to be changed,” she said.

Apps to modify diet and exercise have been designed to use well-established behavior change techniques like setting goals and self-monitoring whether the goals are met.

Spring described a study where participants had multiple common risk factors, including not eating enough fruits and vegetables, eating too much saturated fat, not getting enough physical activity and watching more than 2 hours of television a day.

Rather than asking the participants to work on all four risk factors at once, they asked them to change two – one eating and one activity behavior. Participants were paid $175 to achieve their goals – reaching the half-way point in the first week, the full goal in the second week, and maintaining goal performance for the third week.   Data were uploaded daily and monitored by a coach who provided support and maintained accountability.

“First thing that was shocking to us was that everybody with the exception of one woman was able to meet the behavioral goals without difficulty,” she said. “Most of them did it within a couple of days. No problem.  It didn’t seem to be difficult for them. .”

Then, the participants were just asked to provide data for three days a month for the next six months.  They were no longer paid to meet behavioral goals.  “It shocked us that people didn’t immediately stop doing their healthy lifestyle behaviors,” she said. “They fell off some, but they maintained substantial improvements.

The greatest sustained diet and activity improvements came from the group that was asked to eat more fruits and vegetables and spend less time watching TV. That group started at an average of 3.5 hours of TV, reduced that to 1.5 hours and then inched back up to 2 hours. Their saturated fat consumption also declined: Watching less TV meant less snacking. That group also increased its self-efficacy for changing the diet and activity behaviors.

Conversely, he study found that those who followed the traditional diet of reducing saturated fat and increasing physical activity achieved the least healthy change in diet and activity and also had decreased self-efficacy.

In a follow-up trial, the researchers again asked people to increase fruits and vegetables and decrease leisure screen time, but this time they also asked them to increase physical activity and they didn’t use financial incentives. They observed again that participants made large improvements in diet and activity behaviors and maintained them for 9 months.  And again,   self-efficacy increased for improved fruits and vegetables and limiting sedentary leisure  (TV time), but  self-efficacy did not increase for improving  saturated fat and moderate-vigorous physical activity, even though participants successfully changed those behaviors.

“So what I take from these results is that self-monitoring and changing all behaviors is not equal. Not all behaviors are equally empowering to change,” Spring said. “If we want people to maintain self-monitoring for weight loss despite this disempowering effect on efficacy, I think we’re going to make change easier by providing support through multi-level interventions.”

Psychologists have been very successful at developing behavioral interventions that target the individual to help the person initiate change.  However, Spring said for individuals to effectively maintain that behavior change day in and day out, after the novelty of initiating change has worn off, they need support from a social and cultural context and environment that make the healthy behaviors become the default behaviors.

She cited a study her group did to learn whether giving people a mobile device to help them self-monitor could bolster the effect of socially supportive group treatment to foster weight loss.  The study was done at a Veterans Administration Hospital and participants were in their late 50s or early 60s and unfamiliar with mobile technology.  All participants received the same usual VA weight loss treatment groups, but some self-reported diet and activity on paper and others used an app.

“What we saw was that the app significantly improved weight loss compared to recording on paper.  But what was particularly interesting was how the technology and the socially supportive group context synergized with each other.  People who went steadfastly to the groups but recording on paper didn’t lose any weight. Neither did the people who had the technology but didn’t go to many groups,” she said.  The sweet spot was the people who had the self-monitoring technology and attended groups regularly.  They lost about 15 pounds and maintained it for a year. The technology tools for self-monitoring, combined with the socio-cultural support in the environment was what worked. “

The next step was to determine if technology could be used to deliver social support virtually, in order to overcome access barriers and reduce the need for so many in person group sessions. It seemed like a natural fit, given the success of apps like Facebook.

In a recent obesity trial, Spring treated  participants in half of the usual number of in person group sessions, but to keep the support flowing  also enrolled group members in a chat group, incentivized them as a team for weight loss, and gave them  information about how their teammates  were doing with self-monitoring adherence.

Some groups embraced the idea, but at the end of the day, “it was a bust,” Spring said.

“It seemed like these groups of eight worked well for half of the participants, not so well for the other half,” she said. About half of the participants were interested in knowing how their teammates were doing so they could provide support and receive encouragement themselves. The other half were not interested in engaging with fellow group members and didn’t want to participate in the group chat.

To determine if the small size of the group was an issue, Spring’s group obtained data from 27,000 people in an online weight loss network, and analyzed how their social networks and connectedness in the virtual space related to weight loss.

They found that 89% of enrollees didn’t make a single friend in the online community. But friending was strongly linked to retention in the online treatment program.  About 40% of people who joined the site left after a single logon, and more than half left after about 2 weeks.  But 96% of those who made at least one friend in the virtual community were still participating 9 months in.    “Our challenge is that this was not a randomized experiment,” Spring said. “We did not assign people to make a friend or not.  People assign themselves to the experimental condition of whether they make a friend or not.  If we conclude that friending is helpful for empowering people to make healthy lifestyle changes, how can we get people who are not very extroverted to reach out and befriend somebody else?  Alternatively, how can we grow these connector hubs, who are the folks who welcome people into the group and introduce them to others?”

Today’s treatments are not specifically designed to fit the tempo and context of an individual’s life, Spring said. Current research is attempting to identify ways to tailor interventions for the individual in context. That is where the MD2K Center’s research will help, she added.

“One behavioral problem we’re studying in how to prevent relapse to smoking among people who’ve just recently quit.  Clinically what happens is that if someone has a slip soon after quitting, they’re very likely to go back to smoking. We try very hard to get them to reach out to us for help before they slip, when they sense that a slip is going to happen,” she said. “But they almost never do.  So what we’re trying to do instead is to sense and identify physiological markers to predict those risk states. The one we’re focusing on is stress.  What we’re doing is detecting different levels of stress and micro-randomizing people to receive a stress reduction intervention or no-intervention on these occasions.’

What they hope to arrive at is an evidence-based set of decision rules that tell them when to intervene to prevent or at least delay a smoking lapse.

“It might be, for example, that you need to intervene when somebody is stressed and can benefit from a stress reduction intervention, but on the other hand, it might be that they’re too wound up in their own worries then and can’t process anything else., “ she said. “So maybe they need to be moderately stressed so the intervention is relevant but not overwhelming, or not stressed at all so they can apply their full capacity. That’s what we will know by the end of the experimentation.”

The next few years will dramatically increase our ability to use mobile technology to monitor behavior passively and provide instantaneous feedback.  Those achievements will so reduce the burden and heighten the reward value of self-tracking that I fully expect we’re going to be able to take behavior change intervention to a whole new and exciting level.

You can see Spring's remarks below. Her presentation followed one by Dr. Santosh Kumar, director of the MD2K Center.

 

Copyright © 2017 MD2K. MD2K is supported by the National Institutes of Health Big Data to Knowledge Initiative (Grant #1U54EB020404)

Team: Cornell Tech, GA Tech, Harvard, U. Memphis, Northwestern, Ohio State, Open mHealth, UCLA, UCSD, UCSF, UMass, U. Michigan, Utah, WVU