身体活动依从性 — 数字干预微地图 v0.1
A descriptive micro-map of 10 selected adult digital physical-activity intervention studies, structured around intervention components, behaviour-change mechanisms, time points and maintenance.
This project is jointly maintained by Catherine Ge and Starr Choi. It sits within SISFIT Open Research because it overlaps directly with Catherine's research interests in behaviour change, physical activity and digital health, while remaining methodologically distinct from her independent secondary-analysis projects.
The v0.1 map suggests that the more useful design question is not simply whether an intervention includes tracking, but what happens after measurement and whether the resulting feedback loop stays behaviorally salient during maintenance.
Tracking technologies vary, but monitoring-linked feedback is a common backbone across this set.
Extra SMS, social comparison or prolonged app exposure did not consistently add durable benefit.
Several early between-group differences diminished or disappeared at later time points.
A group can look better because it declines less from baseline.
More walking or leisure-time activity does not automatically imply improved fitness outcomes.
Multi-component trials often cannot identify which component actually produced the observed difference.
| ID | Study | Digital contrast | Key coded result | Source |
|---|---|---|---|---|
| PA001 | Fanning et al. (2017) | Smartphone self-monitoring + factorial goal/points modules | Points-based feedback associated with higher MVPA; all groups increased MVPA. | DOI |
| PA002 | Carr et al. (2013) | Enhanced web intervention vs standard PA website | Advantage at 3 months; no significant between-group difference at 6 months. | DOI |
| PA003 | Zhou et al. (2018) | Adaptive personalized step goals vs fixed 10,000-step goal | Adaptive goals had a better relative trajectory while absolute steps declined in both groups. | DOI |
| PA004 | Slootmaker et al. (2009) | Activity monitor + tailored web advice | No significant PA effect at intervention end or follow-up. | DOI |
| PA005 | Hurling et al. (2007) | Internet/mobile program + real-time activity feedback | Intervention favored moderate PA by accelerometer and self-report. | DOI |
| PA006 | Wang et al. (2015) | Fitbit self-monitoring + SMS vs Fitbit only | Early week-1 SMS advantage did not persist to week 6. | DOI |
| PA007 | Gell & Wadsworth (2015) | Targeted SMS for working women | 12-week group difference was not significant at 24 weeks and largely reflected less decline. | DOI |
| PA008 | Maddison et al. (2015) | Personalized mHealth cardiac rehabilitation | Walking/leisure PA increased; peak VO₂ did not improve vs usual care. | DOI |
| PA009 | Harries et al. (2016) | Personal feedback vs added social comparison | Feedback beat no feedback; social comparison did not significantly beat personal feedback. | DOI |
| PA010 | Fukuoka et al. (2019) | App + counseling acquisition phase; app continuation for maintenance | Initial PA gains were significant; prolonged app use did not improve maintenance vs accelerometer only. | DOI |
The working v0.2 layer extends the map to PA001–PA015 and separates intervention phase, active-comparator structure, adaptation, engagement/exposure, and maintenance interpretation. It is publicly browsable and external methodological validation is still pending.
External feedback question: is our operational mapping of PA006, PA010, PA012 and PA013 onto implementation/exposure versus incremental behavioral-effect failure defensible, and is “preserved behavior” a useful descriptive maintenance outcome category?
v0.2 remains a working layer. Frozen v0.1 and its Zenodo record are unchanged.
This is a convenience micro-map for structural comparison and correction. It is not a systematic review, meta-analysis or intervention-effectiveness ranking. Mechanism labels remain descriptive; BCTTv1 numeric codes are not inferred; null findings are retained; and time point, absolute direction and between-group contrast are kept distinct.
Candidate v0.2 question: Under what conditions does a digital self-monitoring + feedback loop remain behaviorally salient long enough to support physical-activity maintenance?
Co-maintained by Catherine Ge and Starr Choi / SISFIT. Corrections are intended to be transparent and versioned.