Using SMS message and a Mobile Phone Application to Influence Physical Activity in Hispanic Adolescents

Background Wireless technologies offer innovative and adaptive methods to intervene in real-time on health-related behaviors such as physical activity (PA). We have previously shown that a real-time intervention using KNOWME networks, a wireless body-area network with a mobile phone interface, was effective in increasing PA in Hispanic adolescents, a population at high risk for obesity and related diseases[1].

Objectives We previously found that participants’ PA as measured by accelerometer counts was significantly higher after SMS messages from the research team were received, as compared to when no SMS messages were received. These analyses are aimed at discovering if specific message content prompted PA increases, or whether simply receiving any SMS messages was enough to prompt PA increases.

Methods Ten Hispanic adolescents (mean age=16.3±1.7 years, mean BMI percentile=97.2±4.4, 50% female) wore KNOWME Networks and an Actigraph accelerometer over a weekend for a 2.5 day pilot study. KNOWME Networks is a mobile phone application that measures PA through on-body sensors and personalized algorithms. Time-stamped activity data was sent from the application to the research team in real-time, allowing researchers to provide timely reflective feedback and engage participants in SMS message conversations.

The SMS coding scheme was adopted from Motivational Interviewing, an intervention approach that focuses on eliciting and strengthening motivation for behavior change[2]. Messages sent by the researchers were post-hoc categorized as 1) prompting question: question that asked participants what they could do to be physically active when sedentary behavior was observed, 2) affirmation: commending participants for engaging in PA, 3) suggestion: providing suggestions for PA options, 4) neutral: messages that were not related to physical activity. Lagged mixed regression analyses were conducted to determine if different types of reflective messages were associated with an increase in PA ten minutes and twenty minutes after receiving the message.

Results During the 2.5 day pilot study, a mean of 43.1±15.9 total SMS messages, 3.5±2.3 prompting questions, 5.7±3.3 affirmation messages, 1.8±1.9 suggestion messages, and 25.1±5.3 neutral messages were sent to participants. Accelerometer counts were 3332 counts higher in the ten minute period (p<0.001) and 2780 counts higher in the twenty minute period (p<0.01) after a prompting question was sent relative to when no such messages were sent. Accelerometer counts were 3979 counts higher in the ten minute period (p<0.001) and 4065 counts higher in the twenty minute period (p<0.001) after an affirmation message was sent relative to when no messages of this type were sent. No other message types were associated with statistically significant increases in PA.

Conclusions Prompting adolescents to engage in PA using reflective SMS messages based on real-time data is a potentially effective intervention approach. Additional studies with larger sample sizes and theoretically grounded SMS messages are needed to provide further insights into how this methodology can be adapted to promote PA and other health-related behaviors.

References 1. Ogden, C.L., et al., Prevalence of obesity and trends in body mass index among US children and adolescents, 1999-2010. JAMA : the journal of the American Medical Association, 2012. 307(5): p. 483-90. 2. Miller, W.R. and S. Rollnick, Motivational interviewing: Preparing people for change Guilford. New York, 2002.