Animal-Assisted Therapy in Dementia: A Systematic Review and Meta-Analysis of Cognitive, Affective, and Behavioral Outcomes

Abstract

Background/Objectives: Behavioral and psychological symptoms of dementia (BPSD) are common and represent major drivers of caregiver burden and healthcare utilization. Animal-assisted therapy (AAT) is used as a non-pharmacological intervention in dementia care, yet quantitative evidence of its efficacy remains inconsistent. We systematically evaluated the effects of AAT on cognitive, affective, and behavioral outcomes in individuals with dementia. Methods: Following PRISMA 2020 guidelines, we searched seven electronic databases from inception through 27 September 2025 for studies evaluating AAT in dementia. Randomized and non-randomized controlled studies were included in quantitative synthesis. Risk of bias was assessed with the Cochrane Risk of Bias 2 and ROBINS-I tools. Random-effects meta-analyses used standardized mean differences (SMDs), with sensitivity, influence, subgroup, and publication bias analyses conducted where appropriate. Results: Fifty-nine studies met inclusion criteria, including 24 controlled studies with 1881 participants. AAT was associated with reduced overall BPSD severity, although this effect was sensitive to study quality and was not retained after excluding high-risk studies. AAT significantly improved depressive symptoms (SMD, −0.79, 95%CI, −1.33 to −0.26) and cognitive performance (SMD, 1.17; 95%CI, 0.31 to 1.79), with substantial heterogeneity; both effects persisted after excluding studies at high or serious risk of bias. No significant effects were observed for agitation, quality of life, or activities of daily living. Benefits were observed primarily in live animal interventions; robotic subgroups comprised a few studies and were underpowered to detect an effect. Conclusions: AAT may confer meaningful benefits for mood and cognitive engagement in dementia, particularly with live animals, but confidence is limited by methodological heterogeneity and the predominance of studies at moderate to serious risk of bias.