Methods: A quasi-experimental design with two nonequivalent groups (intervention group and comparison group drawn from online forums) and three observations (pretest and two posttests) was used. Data were collected online two weeks before, one to three weeks after, and five to ten weeks after the intervention. The intervention group (n = 25) was 88% female with a mean age of 45.16 years (SD = 10.53) and a mean of 4.74 years (SD = 5.07) since a child’s death. The comparison group (n = 41) was 98% female, with a mean age of 40.10 years (SD = 7.86) and a mean of 3.94 years (SD = 5.33) since a child death.
Mixed-model repeated-measures analyses of variance (ANOVA) were used to assess change over time for both conditions and relative to each other using a significance level of .05 on four self-report outcome measures. The Impact of Event Scale-Revised (IES-R) measured trauma symptoms. It contains intrusion, avoidance, and hyperarousal subscales. The Hopkins Symptom Checklist-25 (HSCL-25) measured depressive and anxious symptoms. The Five Facet Mindfulness Questionnaire (FFMQ) measured mindfulness with five scales: observing, describing, acting with awareness, nonjudging, and nonreactivity. The Self-Compassion Scale-Short Form (SCS-SF) measured self-compassion. All scales showed acceptable to strong internal consistency (α = .69 to .98).
Results: There were significant time by condition interactions with small to medium effect sizes for the IES-R and its three subscales, the HSCL-25 and its depression subscale, and the FFMQ describing, acting with awareness, and nonjudging scales. The intervention group showed significant improvements on all measures of psychological distress at posttest, with gains maintained at follow-up. The comparison group showed significant improvement on the IES-R intrusion subscale at follow-up. The difference between groups was significant at posttest for the IES-R and its intrusion and avoidance subscales and the HSCL-25 anxiety subscale, but only for IES-R avoidance at follow-up. The intervention group showed significant improvements at posttest on the describing and acting with awareness mindfulness scales and self-compassion, although gains were not maintained at follow-up. However, the nonjudge mindfulness facet significantly improved from posttest to follow-up. There were no changes in wellbeing for the comparison group. The difference between groups was significant for acting with awareness at posttest and nonjudging at follow-up.
Conclusions and Implications: Findings suggest the retreat facilitated significant reductions in trauma, depressive, and anxious symptoms that were sustained over time in a sample of bereaved parents. In addition, it temporarily increased the describing and acting with awareness mindfulness facets and increased the nonjudging facet at follow-up. Although further research is needed, this intervention may be an effective way to decrease distress and facilitate adaptive grieving in bereaved parents.