JMIR Publications

ABSTRACT

Background:

Low-income children in the US have high rates of obesity and dental caries. Social media may be an efficient and accessible tool to engage individuals in relevant evidence-based nutrition and dental information and motivate healthier behaviors. Previous research on social media delivered behavioral interventions suggests that engagement may vary by type of posts.

Objective:

To assess (1) engagement with moderator posts in a private Facebook group for caregivers of children ages 1-6 from low-income communities by type of post, and (2) the proportion of caregivers who engaged with moderator posts within a sample two-week period.

Methods:

We enrolled 67 low-income caregivers of children aged 1-6 years old from CT in a private Facebook group. A moderator posted 5 posts per week that included evidence-based information to promote healthy diet and dental health in young children. Posts were polls, informational posts, healthy recipes, and open-ended questions. Informational posts focused on responsive feeding, healthier snacks, sugar sweetened beverages, teeth brushing, and flossing. Polls solicited caregivers’ interest in particular topics, which was used to guide future posts, and were intended to engage caregivers. Information provided was consistent with the 2020 Dietary Guidelines for Americans and post format was informed by the Information Motivation Behavior model. Facebook analytics provided interactions (replies, reactions, poll votes) and impressions on moderator posts from September 2021 to February 2022. We calculated the proportion of caregivers who engaged with moderator posts over a sample 2-week period (2/21/2022-3/4/2022).

Results:

Over 5 months, the moderator posted 110 posts. Polls (N=27) attracted the most engagement, with a median of 5 interactions (IQR: 3-7; range: 0-13) and a median of 24 impressions (IQR: 22-27, range: 20-41) per post. Informational posts (N=38) received a median of 1.5 interactions (IQR: 1-3, range: 0-5) and a median of 22 impressions (IQR: 20-25, range: 10-42) per post. Posts sharing recipes (n=28) had similar engagement with a median of 1 interaction (IQR: 1-2, range: 0-6) and a median of 21 impressions (IQR: 18-24, range: 11-32) per post. Posts with open-ended questions (n=17) had the lowest level of engagement with a median of 0.5 interactions (IQR 0-0.5, range: 0-6) and a median of 19 impressions (IQR: 17.5-24, range: 16-27) per post. During the 2-week sampling period (n=10 posts), 58% of caregivers interacted with at least one moderator post.

Conclusions:

Nearly 6 in 10 caregivers engaged in the group, and non-engaging caregivers may have benefited from involvement despite lack of visible engagement. Polls may be an effective strategy to engage caregivers around child nutrition and dental health than informational posts or posts that ask caregivers open-ended questions about their experiences. Future research should use feedback from caregivers on posts with low engagement to revise posts to better engage caregivers.