Pagoto S. Leveraging social media for health behavior change. Presented at: ObesityWeek Interactive; Nov. 2, 2020 (virtual meeting).

Disclosures: Pagoto reports she does contracted research for Weight Watchers International and is a scientific adviser for Fitbit.
November 12, 2020
5 min read

Meaningful engagement key to weight-loss success in social media intervention programs


Pagoto S. Leveraging social media for health behavior change. Presented at: ObesityWeek Interactive; Nov. 2, 2020 (virtual meeting).

Disclosures: Pagoto reports she does contracted research for Weight Watchers International and is a scientific adviser for Fitbit.
You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

With more people gravitating to social media each year, providers may be able to use platforms such as Facebook and Twitter to create successful, online-only weight-loss intervention programs, according to a presenter.

Sherry Pagoto

During a presentation at ObesityWeek Interactive, Sherry Pagoto, PhD, professor in the department of allied health sciences at the University of Connecticut, said the number of social media users in the U.S. has steadily increased since 2006 across nearly all gender, ethnicity, age and socioeconomic groups. Additionally, a survey of Facebook users conducted by Pagoto and colleagues last summer showed that 69% posted about their health in the past year.

Source: Adobe Stock

“That’s exciting that people are using social media to talk about their health,” Pagoto said. “The downside is there is so much health misinformation out there. If people have this need to use social media to engage about their health, [the misinformation] is a little scary. I began to wonder: Can we create online communities using social media platforms to actually deliver evidence-based lifestyle interventions?”

Pagoto and colleagues have conducted recent trials involving online intervention programs on Twitter and Facebook. Early findings reveal that certain types of engagement may be the key to success for online programs.

Online-only vs. in-person intervention

To assess the potential impact of a social media-based program, researchers conducted a noninferiority, randomized trial in which 329 adults enrolled in a 1-year weight-loss intervention. Participants were randomly assigned to traditional, in-person group intervention meetings (n = 162) or a private Twitter lifestyle intervention group with no in-person visits (n = 167).

Both the in-person and online cohorts were dietitian-led, used a diabetes prevention program lifestyle intervention and included nine to 20 participants in each group. The in-person group had 22 visits lasting 90 minutes. The meetings were held weekly at the start of the program and later transitioned to biweekly and then monthly. The online group was asynchronous, with the intervention taking place through 336 posts on Twitter. The tweet schedule included two posts per day at the start of the program, later transitioning to once-daily posts and then four times per week as the program proceeded. The content from the intervention program was converted into tweets and online articles, with each module divided into 14 posts. On Mondays, participants would post their goals for the week. Fridays would require a weigh-in post, and Sundays would look back at the outcome of the week’s goals. Counselors logged in daily to like posts, reply to discussions and tag participants into new threads.

Researchers compared the mean weight loss and retention between the two groups. Weight was measured at baseline, 6 months and 12 months. Retention was measured through Twitter engagement for the online group and attendance for the in-person group.

After 6 months, the in-person intervention group had a mean weight loss of 4.6% vs. a mean loss of 2.6% for the online group (P = .0032). The in-person cohort also had more participants lose 5% or more weight than the online group (41% vs. 28%; P = .01). There was no significant weight-loss difference between the two groups at 12 months.

The online intervention had slightly better engagement than in-person. Participants on Twitter engaged in a median of 18 modules compared with 14 modules in the in-person group. The online group had 53% of participants engage in the last module vs. 48% in the in-person group, and only 2% of individuals in the online group did not engage at all, whereas 12% of those in the in-person group never showed up to a meeting.

During an acceptability survey conducted after the study, 29% of participants in the online group said the program was time-consuming vs. 44% of those in the in-person groups. However, meaningful engagement was more of an issue online, with more people in the Twitter cohort saying they did not feel connected to the group (27% vs. 16%), did not receive enough diet information (20% vs. 11%) and did not feel comfortable engaging in the group (14% to 6%) compared with participants in the in-person cohort.

When asked about convenience, most participants in both groups said they would have preferred a hybrid program featuring online and in-person elements. Researchers also discovered that three of the nine in-person groups started their own Facebook groups, even though participants were discouraged from engaging with each other online. There was no report of individuals from the online group meeting in-person.

“Inevitably, sometimes this happens, “Pagoto said. “It’s interesting that when people were engaging in-person they did feel this sense that they wanted to be Facebook friends and engage online.”

Superusers and engagement

User engagement has been an issue for many online weight-loss intervention program studies. Pagoto said past studies have seen a wide range of user engagement, from a mean of fewer than one post per participant in one study to an average of more than 100 posts in another. A certain type of user may affect the way a group engages.

“There’s a very small percentage of people in an online patient community that are called superusers,” Pagoto said. “What that means is they engage a ton. These people have been studied, and it has been determined that they are the life of the party in an online social community. They facilitate tie formation between members of the group. They promote retention. They tend to be role models. They are content creators. We’ve been thinking: How can we create such individuals in our weight-loss programs?”

Pagoto and colleagues conducted a pilot feasibility study in which 56 participants were randomly assigned into a superuser group, in which three randomly selected participants were given an incentive to post regularly, or a regular user group with no incentivized superusers. Superusers were paid $25 to post a status update twice a week, reply to the counselor’s post every day and comment when someone else in the group posts. Guidance was provided to the superusers on how to write posts. Researchers measured total engagement by tallying posts, comments and likes from nonsuperusers.

The superuser group had 194 total engagements from nonsuperusers while the regular group had 117 engagements (P = .0361). Most of the additional engagement came from likes; there were 140 in the superuser group vs. 68 in the regular group (P = .048). Researchers found total engagement, likes and comments were predictors for weight loss.

Meaningful engagement

At the end of the study, researchers identified natural superusers who posted just as much as the incentivized superusers. There were six natural superusers in the study who lost a mean of 5.3% of body weight, more than the mean 1.2% weight loss from the three incentivized superusers (P = .0358).

Researchers found most posts from natural superusers were motivational messages or reports about success, whereas incentivized superusers were more likely to provide information or ask for help. After breaking down posts into nine types, researchers found that posts about healthy choices, acknowledgement, weigh-ins, goals or plans, and requests for help were predictors of weight loss, whereas posts about information, support, problems and anything off-topic were not predictors.

“When we are trying to increase engagement in our groups, we want to focus on the types of engagement that we know is reflecting their execution of behavioral strategies,” Pagoto said. “When we design our content, we want to try to elicit this type of engagement.”

Moving forward, researchers plan to look at how to create an online intervention program that elicits the most meaningful engagement and results from participants. This includes further analysis into how to enhance engagement in online programs, whether asynchronous online programs are preferable to synchronous or hybrid interventions, and who responds best to in-person programs vs. online.