“[Abstract] Efficacy of digital health (d-Health) interventions on workers’ physical activity (PA), sedentary behavior, and physiological outcomes remains unclear. This umbrella review searched PubMed, Cochrane Library, and Google Scholar up to October 25, 2024. We identified 24 systematic reviews (SRs) and selected 130 individual studies from these SRs for analysis. The AMSTAR 2 tool rated the quality of most SRs as critically low. Narrative syntheses suggested that d-Health interventions could potentially improve all outcomes compared with no intervention. However, whether d-Health interventions outperform non-d-Health interventions remains uncertain. Meta-analyses showed a significantly small effect of d-Health interventions on step counts, sedentary/sitting time, and weight compared with no intervention, while d-Health interventions slightly improved only moderate-to-vigorous PA compared with non-d-Health interventions. Subgroup analyses identified potential sources of heterogeneity (e.g., risk of bias, control conditions), which may vary between outcomes. Further high-quality studies are needed to evaluate the efficacy of d-Health interventions.
[Discussion] To the best of our knowledge, this is the first umbrella review to summarize the efficacy of d-Health interventions on PA [physical activity]-related, SB [sedentary behavior]-related, and physiological outcomes in workers. We included 24 SRs and 130 individual studies. Low to very low-quality evidence suggested that d-Health interventions had small but significant effects on PA-related, SB-related, and physiological outcomes compared to no intervention. However, there was no clear evidence supporting the superiority of d-Health interventions over non-d-Health interventions, such as exercise sessions and face-to-face counseling. Additionally, insufficient evidence was available to determine whether one type of d-Health intervention was more effective than another.
[..] d-Health intervention led to a small but significant weight reduction compared with no intervention (SMD [standardized mean difference]: −0.24; 95CI: −0.41 to −0.07; I2 = 75.2%); while no significant weight change was observed compared with non-d-Health intervention (SMD: −0.01; 95% CI: −0.06 to 0.04; I2 = 0%). Although previous SRs [systemic reviews] showed inconsistent results24,27,28, our umbrella review supports the efficacy of d-Health interventions in improving physiological outcomes compared with no intervention. However, we did not find clear evidence regarding whether d-Health interventions outperform non-d-Health interventions or which type of d-Health intervention is most effective. Further research is needed to obtain conclusive evidence.”
Full article, M Iwakura, C Ozeki, S Jung et al., NPJ Digital Medicine, 2025.4.14