Keltz (2025)
Secondary analysis of 3-arm RCT (Simon 2021, NCT01912287); logistic regression; N=190 subsample; 2-site; homework engagement as predictor of response
Homework Engagement Analysis from KY vs CBT vs SE Trial (Simon 2021 — Secondary Analysis) PARENT TRIAL: Simon et al. (2021) — 226 adults with GAD randomized to KY (n=93), CBT (n=90), or Stress Education (n=43). SUBSAMPLE: N=190 who attended ≥2 sessions and submitted ≥1 homework log. KY HOMEWORK: Audio-guided sequence of movements coordinated with breath, followed by timed breathing with mindful awareness. Same audio recording used for each session. Prescribed 20 min/day. CBT HOMEWORK: Activities and worksheets to practice session skills. Prescribed 20 min/day. SE HOMEWORK: Audio recordings about nutrition and lifestyle. Prescribed 20 min/day. HOMEWORK MEASURES: (1) Staff-rated compliance scale (0-6 Likert), (2) Self-reported days/week from written logs. KEY FINDING: Staff-rated homework compliance predicted response for CBT (OR=2.49, p=.005) and KY (OR=1.66, p=.049), but not SE. KY and CBT showed greater overall homework engagement than SE. DROPOUT: 34% (23% CBT, 26% KY, 49% SE) dropped out prior to session 12, on average after session 10.