JMIR Publications

ABSTRACT

Background:

Physical inactivity is a major public health challenge, particularly among older adults. Digital behavioral change interventions (DBCIs) offer scalable approaches, but evidence on their effectiveness, sustained engagement, and equity remains fragmented.

Objective:

To systematically collect and synthesize evidence on (1) the efficacy of DBCIs for increasing physical activity and reducing sedentary behavior in older adults, (2) adherence and retention patterns, and (3) how included studies addressed the digital divide.

Methods:

A comprehensive search of PubMed, Web of Science, PsycINFO, Cochrane Central, and Scopus was conducted on studies published between 2016 and 2026. Eligible studies evaluated DBCIs aimed at increasing physical activity in community-dwelling adults aged 50 years. Reviewers independently screened records, extracted data, and assessed risk of bias using RoB 2.0 for randomized trials and ROBINS-I for non-randomized studies; a second reviewer verified data and approved the selection. The certainty of evidence was assessed using the GRADE framework.

Results:

Twenty-three studies were included, comprising 21 randomized controlled trials and 2 non-randomized studies, with 2,893 participants overall; two ongoing protocol studies were also identified but did not contribute outcome data. Owing to substantial heterogeneity in interventions, outcomes, and reporting, findings were synthesized. In terms of efficacy, DBCIs consistently improved daily step count, with reported improvements ranging from +797 to +2,190 steps/day; 9 studies, N =790, GRADE: MODERATE). Moderate-to-vigorous physical activity also generally increased (+6.5 to +78.6 min/week; 7 studies, N =616, GRADE: MODERATE ), while total physical activity increased by approximately +22 to +209 min/week; 6 studies, N =586, GRADE: MODERATE). Evidence for sedentary time was less consistent (GRADE: LOW). Mean adherence across eight studies was 81% (range 65-97%; 8 studies), and mean retention across studies was 87% (range 70%-98.7%). In terms of adherence and retention patterns, interventions incorporating interpersonal strategies, including coaching, peer interaction, or group-based support, appeared to achieve stronger and more sustained engagement than self-directed digital approaches. In relation to digital divide, exploratory evidence suggested that digital divide factors remained evident: lower computer literacy, older age, lower education, and reduced social support were associated with lower engagement, with reported odds ratios ranging from 0.61 to 5.39; however, these factors were not prespecified primary outcomes in most studies. Among the 23 studies with outcome data, 6 (26%) were rated low risk of bias, 10 (44%) had some concerns, and 7 (30%) were rated high risk.

Conclusions:

DBCIs appear effective for increasing physical activity in older adults, particularly step count and overall activity, and engagement is generally favorable when interventions include interpersonal support. However, digital divide-related barriers continue to shape participation and sustained use. Future studies should focus on uniform reporting of outcomes, longer follow-up, and the integration of digital interventions with relational support into integrated intervention designs. Clinical Trial: PROSPERO (Registration no-CRD420261467897)