Exploring the Mechanisms of Digital Health Technology-Driven Exercise During Pregnancy: A Realist Review

Exercise during pregnancy is a key factor in promoting the health of both mother and baby, yet adherence to exercise programs among pregnant women worldwide is generally low. Although digital health technologies have been used in exercise interventions during pregnancy and have shown initial promise, the mechanisms by which they encourage pregnant women to participate in exercise remain unclear.

This study aims to explore the mechanisms by which digital health technologies promote physical activity during pregnancy to guide the design of relevant intervention programs.

This study adopted a realist review approach, conducted in three iterative stages: (1) development of an initial programme theory; (2) systematic screening of published research and grey literature against pre-defined inclusion criteria, including original research from seven electronic databases (CNKI, VIP Database, PubMed, Cochrane Library, Web of Science, Embase, and ScienceDirect) searched from their inception to April 2026; and (3) analysis, extraction, and synthesis of evidence from included studies to refine context-mechanism-outcome (CMO) configurations and draw conclusions.

We included 15 studies, and the analysis identified six positive mechanisms and two negative inhibitory mechanisms through which digital health interventions influence exercise during pregnancy. The positive mechanisms promote exercise behavior among pregnant women and enhance exercise adherence in multiple ways; the negative mechanisms, however, undermine the positive mechanisms and limit the effectiveness of the interventions. Specific contextual factors moderate the activation or inhibition of these mechanisms, and the combined effects of positive and negative mechanisms jointly determine the intervention's outcome.

This study has identified key mechanisms by which digital health technologies promote physical activity during pregnancy, providing a theoretical basis for clinical interventions and improving participation rates and adherence to exercise programs among pregnant women.

Only published studies were included, which may introduce publication bias. The variability in study contexts and populations limits the generalizability of the identified mechanisms.