Stimulation-Oriented Interventions for Behavioral Problems among People with Dementia: A Systematic Review and Meta-Analysis

PURPOSE: This study was a systematic review and meta-analysis designed to investigate the effects of stimulation-oriented interventions for behavioral problems among people with dementia. METHODS: Based on the guidelines of the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA), a literature search was conducted using seven electronic databases, gray literature, and other sources. Methodological quality was assessed using the Scottish Intercollegiate Guidelines Network (SIGN) for randomized controlled trials (RCTs). Data were analyzed using R with the 'meta' package and the Comprehensive Meta-Analysis (CMA 2.0) program. RESULTS: Sixteen studies were included for meta-analysis to investigate the effect of stimulation-oriented interventions. The quality of individual studies was rated as '++' for eight studies and '+' for the rest. The effect sizes were analyzed according to three subgroups of interventions (light, music, and others); Hedges' g=0.04 (95% CI: -0.38~0.46), -0.23 (95% CI: -0.56~0.10), -0.34 (95% CI: -0.34~0.00), respectively. To explore the possible causes of heterogeneity (I²=62.8%), meta-regression was conducted with covariates of sample size, number of sessions, and length of session (time). No moderating effects were found for sample size or number of sessions, but session time showed a significant effect (Z=1.96, 95% CI: 0.00~0.01). Finally, a funnel plot along with Egger's regression test was performed to check for publication bias, but no significant bias was detected. CONCLUSION: Based on these findings, stimulation-oriented interventions seem to have a small effect for behavioral problems among people with dementia. Further research is needed to identify optimum time of the interventions for behavioral problems among dementia pateints.