Mobile-delivered yoga linked to greater reduction in depressive symptoms than self-care control

Seventy-one participants (94.7%) completed at least 9 of the 12 sessions and were analyzed. Both intervention groups used motion-detecting technology to guide practice, with no live therapist involvement. The primary outcomes were perceived stress (Perceived Stress Scale) and depressive symptoms (Patient Health Questionnaire-9), measured at baseline, immediately after the 12-session program, and 4 weeks later. For depressive symptoms, the yoga group showed significantly greater improvement than control at 4-week follow-up (β=−3.67; P<0.001), representing a clinically meaningful reduction. Perceived stress decreased across all groups over time (P<0.001), but the between-group difference was not significant. For cognitive stress responses, yoga showed a consistent linear improvement while exercise showed immediate but less sustained gains. The evidence is moderately strong (RCT design, concealed allocation, in-person assessments), though limited by lack of blinding and a relatively small sample over a short timeframe.

The yoga group's advantage over control emerged most clearly at the 4-week follow-up, not immediately post-intervention, suggesting a delayed or cumulative effect rather than acute symptom relief. This contrasts with the exercise group, which showed early gains that did not sustain—an important distinction for understanding how different movement practices affect mood regulation. Cognitive stress responses (measured on a scale capturing rumination, worry, and intrusive thoughts) showed differential patterns: yoga produced steady, linear improvement while exercise produced a sharp initial dip followed by plateau. Heart rate variability and EEG data were collected as physiological markers but were not detailed in the results presented, leaving open whether changes in autonomic nervous system activity or brain oscillations explained the psychological improvements. The control group (mobile CBT-based self-care education with no movement) also improved significantly over time, indicating that digital health interventions and time-based expectancy effects matter; the yoga advantage was measured against an active control, not against no treatment. The 12-session, fully automated design (no therapist contact) is practically valuable for workplace scaling but means individual tailoring and real-time feedback were minimal. Longer follow-up and larger samples are needed to assess whether benefits persist beyond 4 weeks and whether effects are specific to yoga or generalizable to other structured, motion-guided practices.