Exercise & Training

Automated Physical Activity Support for Adults and Youth From Low-Income Communities: Single-Arm Pilot Study.

TL;DR

Personalized mHealth physical activity interventions integrating wearable step trackers with automated text messaging appear to be feasible and acceptable among adults and youth from low-income communities, with step-count findings showing promise particularly for individuals further from meeting physical activity guidelines.

Key Findings

Enrollment rates were approximately 64% for adults and 63% for youth recruited through community health events in low-income urban communities.

  • 84 of 132 eligible adults enrolled (64%)
  • 31 of 49 eligible youth enrolled (63%)
  • Participants were recruited through community health events held in low-income urban communities
  • Final sample included 83 adults and 31 youth

Retention for physical activity measures was 84% for adults and 77% for youth over the 7-week intervention.

  • 70 of 83 adults were retained for physical activity measures (84%)
  • 24 of 31 youth were retained for physical activity measures (77%)
  • The intervention was a single-arm pre-post design lasting 7 weeks
  • Retention was measured specifically for physical activity outcome data

Message delivery fidelity was extremely high, with 99% of intended automated text messages successfully delivered.

  • 3,910 of 3,955 intended messages were delivered (99%)
  • Messages were sent 4 days per week
  • Messages focused on goal setting, self-monitoring, reinforcement, contextual factors, and self-efficacy
  • Most messages were personalized using step-count and step-goal data, branching logic, and 2-way question-and-response formats

Adults and youth showed high adherence to wearing the Garmin activity monitor throughout the study.

  • Adults wore the Garmin on 82% (45/56) of study days
  • Youth wore the Garmin on 79% (44/56) of study days
  • Study duration was 56 days (approximately 7 weeks)
  • Wearing adherence was tracked as a feasibility measure

Overall acceptability ratings were high, ranging from 83% to 100%, with nearly all participants indicating they would recommend the program.

  • 97% of adults (75/77) indicated they would recommend the program
  • 100% of youth (27/27) indicated they would recommend the program
  • Overall acceptability ratings ranged from 83% to 100% across survey items
  • Acceptability was measured through survey items and engagement with 2-way messages

Adults and youth replied to a mean of 2.6 and 3.2 two-way text messages respectively, with higher engagement among adults participating with their child.

  • Adults replied to a mean of 2.6 (SD 2.2) of 7 messages that requested a reply
  • Youth replied to a mean of 3.2 (SD 2.7) of 7 messages that requested a reply
  • Higher engagement was observed among adults who participated with their child (parent-child dyads)
  • Two-way message response rate served as an acceptability/engagement measure

Pre-post changes in daily steps were positive but not statistically significant for either adults or youth in the overall sample.

  • Adults showed a change of β=240 steps/day (95% CI -387 to 866)
  • Youth showed a change of β=413 steps/day (95% CI -877 to 1703)
  • Both confidence intervals crossed zero, indicating non-significant changes
  • Changes were analyzed using mixed-effects linear regression

Participants in the highest tertile of engagement showed larger step-count increases than the overall sample.

  • Adults in the highest engagement tertile showed β=584 steps/day (95% CI -784 to 1952; n=19)
  • Youth in the highest engagement tertile showed β=941 steps/day (95% CI -827 to 2709; n=11)
  • These changes remained statistically non-significant (confidence intervals crossed zero)
  • Engagement was defined based on response to 2-way messages

Participants who were meeting less than two-thirds of physical activity guidelines at baseline showed larger step-count increases, with adults approaching statistical significance.

  • Adults meeting less than two-thirds of physical activity guidelines at baseline showed β=609 steps/day (95% CI -30 to 1247; n=47)
  • Youth in the same subgroup showed β=1406 steps/day (95% CI -94 to 2907; n=22)
  • The adult subgroup confidence interval nearly excluded zero (-30 to 1247), suggesting a trend toward significance
  • The youth subgroup also had a confidence interval nearly excluding zero (-94 to 2907)

The intervention integrated a Garmin activity monitor with automated, theory-based, personalized text messages using behavior change techniques.

  • Messages were sent 4 days per week over 7 weeks
  • Behavior change techniques included goal setting, self-monitoring, reinforcement, contextual factors, and self-efficacy
  • Personalization included calculations based on step-count and step-goal data, branching logic, and 2-way question-and-response messages
  • The intervention was described as a 'beta-version' mHealth program
  • ClinicalTrials.gov registration: NCT05110508

What This Means

This research tested a 7-week mobile health program designed to help adults and youth from low-income urban communities become more physically active. Participants received a Garmin fitness tracker and automated, personalized text messages four days per week. The messages used evidence-based strategies like goal-setting, self-monitoring, and encouragement, and many messages were customized based on each person's actual step counts and goals. The study enrolled 83 adults and 31 youth through community health events and tracked whether the program was practical to deliver and whether participants liked it. The study found strong evidence that this type of program is both feasible and well-received in this population. Nearly all messages were delivered as intended (99%), participants wore their fitness trackers on about 80% of study days, and 97-100% of participants said they would recommend the program to others. While overall increases in daily steps were positive (about 240 more steps per day for adults and 413 for youth), these changes were not statistically significant in the full group. However, participants who were least active at the start of the study and those who engaged most with the program showed larger step increases, suggesting the program may be especially beneficial for those with the most room for improvement. This research suggests that wearable-integrated, automated text message programs are a promising approach for supporting physical activity in low-income families, a group that has been understudied in mobile health research. The findings are particularly encouraging for reaching individuals who are furthest from meeting physical activity guidelines. The authors note that larger, more rigorous studies—including randomized controlled trials—are needed to fully evaluate how well this type of program works, and that parent-child participation together may be worth exploring further.

Check Your Own Numbers

Upload your bloodwork. We'll cross-reference your results against this study and 4,700 others.

Upload Your Labs

Have a question about this study?

Citation

Carlson J, Materia F, Moon M, Ryu S, Yeager C, Steel C, et al.. (2026). Automated Physical Activity Support for Adults and Youth From Low-Income Communities: Single-Arm Pilot Study.. JMIR mHealth and uHealth. https://doi.org/10.2196/76991