Can Layered Support Get Rural Adults Moving More?

A modest win at the 150-minute line

Layered support can help rural adults stay active, but Heartland Moves did not produce evidence of a broad increase in weekly exercise volume. After 12 months, 49.0% of intervention completers met aerobic activity guidelines, compared with 40.6% of control completers: an 8.4-percentage-point difference. Cluster-adjusted analysis estimated nearly double the odds of meeting the guideline (adjusted odds ratio 1.99; 95% CI 1.01–3.93; p=0.045).

That is a meaningful result, particularly in communities often left out of digital-health research. It is also a narrow one. The intervention did not significantly improve reported recreational, occupational, transportation, or total moderate-to-vigorous physical activity (MVPA) minutes. The most accurate reading is that the package helped more follow-up participants remain above the 150-minute aerobic threshold during a difficult year; it did not demonstrate a large rise in total activity across the group.

The evidence comes from Heartland Moves, a 2025 cluster-randomized trial in southeast Missouri. It asked a more useful public-health question than whether a phone app works under ideal conditions: can a community deliver several forms of support at once, and can that support change behavior over a year?

A real-world rural trial, interrupted by reality

Researchers pair-matched 14 rural Missouri communities by population size, poverty, and percentage of non-White residents, then assigned communities to intervention or control conditions through one of 64 allocation permutations. The towns ranged from 1,800 to 16,000 residents and covered Rural-Urban Continuum Codes 3 through 9.

The planned intervention combined mobile messages with walking groups, trail events, printed guides, community outreach, and local environmental support. COVID-19 changed what participants actually received. The SMS program largely continued: participants could opt into a 24-week exercise-text program delivering three messages each week, plus community-tailored outreach twice weekly for 12 weeks. In-person walking groups, leader tracking, and trail events were sharply reduced. Guides arrived by mail, communities received mailers, and one community added two billboards.

Of 1,241 people assessed at baseline, only 492 completed follow-up, about 40%. Among these completers, the mean age was 58.0 years; 67.9% were women; 71.7% had a high school education or less; 57.0% reported household income below $50,000; and 16.1% were non-White. The study therefore speaks directly to lower-income rural adults, while offering limited insight into more racially diverse rural populations.

The positive endpoint and the missing one

The trial's positive result concerns guideline attainment, a binary outcome. The crude 8.4-point difference suggests roughly one additional person meeting the guideline for every 12 people followed, if the observed contrast reflects a causal effect. Since participants entered the study with relatively high guideline attainment—51.0% in intervention completers and 53.0% in control completers—the result may partly reflect prevention of decline rather than recruitment of many inactive people into regular exercise.

Continuous activity outcomes tell a less persuasive story. Adjusted between-group differences at follow-up were 57.8 more recreational MVPA minutes per week (p=0.233), 14.2 more occupational minutes (p=0.897), 15.7 more transportation minutes (p=0.728), and 92.9 more total MVPA minutes (p=0.563). None was statistically significant.

This discrepancy can happen when many people sit near a threshold, especially when activity comes from self-report. A person who reports a small increase may cross the 150-minute boundary even if average minutes across the whole sample barely change. That does not erase the guideline-attainment result. It does limit what it can prove.

Why SMS looks plausible, but remains unproven

Participants who received at least half of the exercise texts reported higher adjusted total MVPA than those who received none. That finding makes text support a plausible contributor, not the established active ingredient. SMS participation was optional, and people who chose or continued to receive messages may have differed in motivation, access, health, or baseline habits. Heartland Moves did not randomly assign participants to individual intervention components.

Broader evidence supports modest expectations for mobile support. A 2021 meta-analysis of 117 randomized mHealth trials, involving 21,118 participants, found an end-of-intervention MVPA effect of SMD 0.28 (95% CI 0.21–0.35) and a smaller long-term effect of SMD 0.19 (95% CI 0.11–0.27). The review also found high risk of bias in 80.3% of included studies and rated much of the evidence low or very low certainty.

A separate meta-analysis of smartphone apps and activity trackers pooled 28 studies with 7,454 adults and estimated an SMD of 0.350 (95% CI 0.236–0.465), which the authors translated to roughly 1,850 additional daily steps. Text messaging and personalization performed better in that review. Those results support the idea that repeated prompts can help, but they do not identify SMS as the cause of Heartland Moves' result.

Stronger delivery has produced stronger signals

Other trials suggest that social and community structure may add value when programs can sustain them. In a cluster trial of 72 rural older adults, peer support plus a mobile-app walking program increased daily steps by 408 relative to control after three months. That is an incremental gain, but it shows that rural adults can respond to combined digital and social support.

The larger PEP4PA cluster trial enrolled 476 low-income older adults and increased objectively measured MVPA by about 10 minutes per day over 24 months. Its intervention maintained peer leaders, twice-weekly walks, coaching, self-monitoring, and center-level support. It used device-based measurement, which avoids some of the uncertainty of recall-based activity questionnaires.

These studies do not make Heartland Moves a failed version of PEP4PA. They used different populations, intervention intensity, measurement methods, and follow-up periods. They do show why COVID-related loss of interpersonal and community contact matters: the published Heartland Moves result reflects the program that communities could deliver during the pandemic, not the full program researchers originally intended.

Mechanism evidence remains suggestive. An exploratory study in rural women found that changes in social support and self-efficacy helped explain physical-activity and healthy-eating pathways in a multilevel intervention. Mediation analysis cannot prove that either factor caused the change, but it gives a reasonable rationale for including social support rather than treating movement as a private willpower contest.

The attrition problem

Attrition is the biggest reason to restrain confidence. Follow-up included only 492 of 1,241 baseline participants, and completers differed from non-completers in education and recreational physical activity. The study did not report an imputation-based intention-to-treat correction for missing outcomes. Selective retention could therefore account for part of the apparent benefit.

Measurement adds another constraint. COVID forced the study to abandon device-based activity measurement and rely on the Global Physical Activity Questionnaire. Reported domain minutes were highly skewed and often zero. Self-reported activity can capture perceived behavior change, but it is less precise than accelerometry and vulnerable to recall and reporting error.

Longer-term prevention trials also argue against grand claims. The 48-month PROPELS trial tested group walking support with pedometers, with or without tailored mHealth support, among people at high risk of diabetes. It fits a familiar pattern: behavioral support can increase or sustain activity, but it does not reliably create dramatic long-term change. The InterWalk trial likewise found that app-supported lifestyle care could increase or sustain physical activity in adults with type 2 diabetes, within the bounded effects typical of real-world delivery.

Verdict

Heartland Moves offers credible evidence that a layered rural-community program can improve the odds that follow-up participants meet aerobic activity guidelines. The 8.4-point absolute difference matters, especially because the intervention reached lower-income rural communities during COVID disruption.

It does not establish that the program substantially increased total activity minutes, that SMS caused the benefit, or that the original multilevel design would have performed the same way without pandemic constraints. For communities, the practical lesson is modest but useful: personalized digital contact can supply scalable support, while peer and local infrastructure may produce more when they are delivered consistently. A better replication would retain more participants, use device-based activity measurement, preserve community programming, and randomize the individual components that researchers want to credit.