|Do animal-assisted activities effectively treat depression: a meta-analysis
|Souter MA, Miller MD
This review found some empirical support for the therapeutic effectiveness of dog-assisted activities/therapy for treating depression. The authors' conclusions should be regarded with caution, given the poor reporting of review methods, the lack of quality assessment and the paucity of primary data.
To assess the effectiveness of animal-assisted activities and animal-assisted therapy for reducing depressive symptoms in humans.
The following databases were searched: PsycINFO, ERIC (Education Resources Information Center), Social Services, Sociological Abstracts, Academic Search Premier, NIMH (National Institute of Health database), Health and Wellness Resource Centre, MEDLINE, AGRICOLA, Melvyl Catalog, CINAHL, Health Source, Kluwer, Project MUSE, ScienceDirect, Cochrane Database of Systematic Reviews and Dissertation Abstracts International. Search terms were reported. The journal 'Society and Animals', relevant websites on animal-assisted activities/therapy and of universities with programs of research into human-animal relationships were handsearched. Reference lists of retrieved papers were also scanned. Only papers published in English were considered, but no date restrictions were applied.
Studies were only eligible if they used a randomised design to compare animal-assisted activities or animal-assisted therapy with a control group, and reported depressive symptoms via a self-report questionnaire. Studies also had to report sufficient details of the results to enable calculation of effect sizes. Studies of pet ownership were excluded.
Most of the included trials (all RCTs) evaluated animal-assisted activities compared to no treatment or an alternative control. The animal-assisted activity/therapy intervention lasted between eight to 68 minutes, and total number of visits ranged between one and 24 visits. Four trials used individual animal visits (dogs), one trial used group visits and also included a cat. Depressive outcomes were reported on validated measures including Zung Self Rating Depression Scale; Beck Depression Inventory (BDI) and BDI II; Geriatric Depression Scale; and the NIMH Mood Scales. All included trials were conducted in the USA between 1984 and 2000, settings included hospital units, nursing homes and psychiatric wards. Participants' mean age ranged from 47 to 85 years, with all but one trial including females (50 to 95% female).
The authors did not report how many reviewers performed the study selection.
Assessment of study quality
The authors did not state that they assessed validity.
Where trials reported multiple comparison arms, data were extracted from all arms, but only the no-treatment/activity arm was used for calculations. Cohen's d or the effect size for each trial was calculated. The calculation was modified to take into account the direction of the outcome measure indicating improvements, and adjusted for small sample size where necessary.
The authors did not report how many reviewers performed the data extraction.
Methods of synthesis
A random-effects model was used to calculate standardised mean differences (SMD) and associated 95% confidence intervals (CI) based on the corrected effect sizes for each individual trial. Heterogeneity was assessed using the Q statistic and I2. Publication bias was quantified using the fail-safe N calculation.
Results of the review
Five RCTs were included in this review (n=254 patients, n=182 included in the analyses). Sample sizes were small overall (15 to 80 patients) and ranged from five to 25 patients per treatment group.
Meta-analysis: Meta-analyses of all 5 trials indicated a beneficial effect of animal visits over control (SMD 0.61, 95% CI 0.03 to 1.19), but significant heterogeneity was noted (Q=13.61, I2=70.61%). The forest plot showed that two trials reported a statistically significant difference from animal visits, two trials showed no significant difference, and one trial found a non-significant improvement for the control group. When the latter trial was excluded, similar results were found (revised SMD 0.87, 95% CI 0.45 to 1.28) without significant heterogeneity. The authors discussed some possible reasons for the conflicting results of this trial.
Fail-safe N: To reduce the mean standardised mean difference to 0.50, one unpublished trial with an effect size of 0 would be required. To produce a standardised mean difference of 0.2, ten additional trials with effect sizes of 0 would be required.
The results offered some empirical support for the therapeutic effectiveness of dog-assisted activities/therapy for treating depression, but further well designed studies, focusing on physiological measures, would be beneficial.
This review addressed a clear question with well specified inclusion and exclusion criteria. The searches were comprehensive, and although they did not consider non-English language papers, unpublished studies were sought and included. The methods used in the review were poorly reported, so it was difficult to exclude reviewer bias/error. Also, no formal quality assessment was carried out, so the reliability of the primary data was unclear. The analysis appeared to have been appropriate and heterogeneity was assessed and explored. The authors' conclusions should be regarded with caution, given poor reporting of review methods, lack of quality assessment and the limited evidence available from few trials with small treatment groups.
Implications of the review for practice and research
Practice: The authors did not state any implications for practice.
Research: The authors stated that further research in the area of animal-assisted activities/therapy should consider physiological outcome measures, such as heart rate and blood pressure, which serve as useful health risk markers. Research should also address the degree to which the positive effects of animal-assisted activities/therapy are attributable to contact with the person facilitating the animal interaction, and more generally trials should include longer follow-up periods.
Souter MA, Miller MD. Do animal-assisted activities effectively treat depression: a meta-analysis. Anthrozoos 2007; 20(2): 167-180
Subject indexing assigned by CRD
Animals, Domestic; Bonding, Human-Pet; Depression; Humans
Date bibliographic record published
Date abstract record published
This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.