Abstract
Animal-assisted therapy (AAT) may affect health via an increase in perceived social support and social interaction. A meta-analysis was conducted to determine its effects on selected populations with poor social functioning. Studies involving elderly participants and those with depression and schizophrenia were searched in Medline, PsycINFO and Cochrane up to January 2009. Matched or controlled trials incorporating pre- and post-test outcome measures and with at least five participants subjected to a multiple-day AAT intervention were selected (n=21). Participants' characteristics, study design, intervention features, outcome variables, reported effects and study quality were retrieved. Design-specific effect sizes were pooled using random-effects models. Heterogeneity of effects by study characteristics, including target population and intervention features, was explored through meta-regression. Pooled effect sizes were calculated for social functioning, depression, anxiety, behavioural disturbances, loneliness, daily living skills and cognitive status. As hypothesised, AAT improved social functioning (pooled effect size = 1.06, n=275). Moderate effects were found for depression (−0.34, n=447), anxiety (−0.29, n=291) and behavioural disturbances (−0.32, n=367). Effects on social functioning and depression were larger in individuals with psychiatric conditions while behavioural disturbances were reduced in patients with dementia. The inconsistent methodological characteristics of the studies meta-analysed suggest a conservative interpretation of these findings.
Acknowledgements
We thank Dr. Stephen N. Haynes for his insightful comments on previous versions of this manuscript. Dr. Anthony L. Podberscek provided reprints of some of the meta-analyzed studies. We also thank Ana Lungu for translating studies in Russian during the screening of references.