Comparative Efficacy of Spirituality, Cognitive, and Emotional Support Groups for Treating Eating Disorder Inpatients

Abstract

Spiritual interventions are rarely used in contemporary treatment programs and little empirical evidence is available concerning their effectiveness. The purpose of the present study was to evaluate the effectiveness of a spiritual group intervention for eating disorder inpatients. We compared the effectiveness of a Spirituality group with Cognitive and Emotional Support groups using a randomized, control group design. Participants were 122 women receiving inpatient eating disorder treatment. Patients in the Spirituality group tended to score significantly lower on psychological disturbance and eating disorder symptoms at the conclusion of treatment compared to patients in the other groups, and higher on spiritual well-being. On weekly outcome measures, patients in the Spirituality group improved significantly more quickly during the first four weeks of treatment. This study provides preliminary evidence that attending to eating disorder patients' spiritual growth and well-being during inpatient treatment may help reduce depression and anxiety, relationship distress, social role conflict, and eating disorder symptoms.

This research was supported by the John Templeton Foundation, Radnor, Pennsylvania.

An earlier version of this article was presented at the annual convention of the American Psychological Association on August 22, 2001, in San Francisco, California.

Notes

¹ We set our critical value for rejection of the null hypothesis (no differences between the treatment groups) at alpha < .05. Given that we conducted eight separate ANCOVAs, there is a 40% chance that one of the significant findings reported above occurred by chance. Performing a Bonferroni correction would mean setting the critical value at an alpha of .00625, which would have led to the conclusion that the only statistically significant differences in Table 1 were on the OQ-45 relationship distress and social role conflict subscales. Given that research is in an early stage in this domain, and given the relatively small sample sizes in our treatment groups and the relatively low statistical power of the ANCOVAs, we felt that a critical value of .00625 was overly conservative. We preferred the possibility of making a Type I error rather than a Type II error in this situation.