RESUMEN
This study examined self-reported and actigraphy-assessed sleep and depression as moderators of the effect of a Tibetan yoga intervention on sleep and depression among women undergoing chemotherapy for breast cancer. This is a secondary analysis of an RCT examining a 4-session Tibetan yoga program (TYP; n = 74) versus stretching program (STP; n = 68) or usual care (UC; n = 85) on self-reported sleep (Pittsburgh Sleep Quality Index (PSQI), actigraphy-assessed sleep efficiency (SE)) and depression (Centers for Epidemiological Studies Depression Scale; CES-D) for women undergoing chemotherapy for breast cancer. Data were collected at baseline and 1-week and 3-month post-intervention. Baseline PSQI, actigraphy-SE, and CES-D were examined as moderators of the effect of group on PSQI, actigraphy-SE, and CES-D 1 week and 3 months after treatment. There was a significant baseline actigraphy-SE × group effect on PSQI at 1 week (p < .001) and 3 months (p = .002) and on CES-D at 3 months (p = .049). Specifically, the negative association of baseline actigraphy-SE with subsequent PSQI and CES-D was buffered for women in the TYP and, to a lesser extent in STP, compared to those in the UC. Baseline PSQI and CES-D were not significant moderators of the effect of group on any outcome. Behaviorally assessed sleep may be a more robust indicator of which patients are most appropriate for a yoga intervention than self-reported sleep quality. Women with poor sleep efficiency may derive the greatest benefit in terms of sleep quality and mood from a yoga intervention.