A systematic review & meta-analysis of randomised controlled cognitive-based interventions for dementia – restorative, compensatory, & mixed approaches

Background: Cognitive impairments are a central feature of dementia and have a significant impact on those with the disease and their caregivers. Cognitive-based interventions for dementia, (e.g., stimulation, training, and rehabilitation), are often used to ameliorate the symptoms of cognitive impairments. These interventions are based on processes that restore or compensate for the impairments found in dementia. The aim of this systematic review and meta-analysis is to examine the existing research and establish the most effective approach for improving cognitive and functional outcomes for persons with dementia (PwD). Objectives: To evaluate and compare the efficacy of restorative and compensatory interventions to improve cognitive and functional outcomes for PwD. Search Methods: PsycInfo, PubMed, SCOPUS, and CENTRAL were searched between 01.01.2000 and 31.12.2015 in January 2016. Randomised controlled trials (RCTs) of restorative, compensatory, and mixed interventions, published in English, were considered for inclusion. Data Collection & Analysis: 24 RCTs were included in the systematic review and meta-analysis – 18 restorative, two compensatory, and four mixed. The unit of analysis was change scores from baseline to post-intervention for all groups. Estimates of treatment effect were calculated using a random effects model. Statistical heterogeneity was measured using a Chi2 statistic. Where heterogeneity was found a moderator analysis was conducted. Main Results: Restorative interventions demonstrated positive effects on cognition, language, and working memory. Compensatory approaches led to improvements on PwD and caregiver quality of life. Mixed approaches increased self-efficacy for PwD. Moderator analysis indicated that restorative interventions led to significantly greater improvements on language compared to mixed. Conclusions: Restorative interventions demonstrated greater improvements on measures of language. A lack of RCTs on compensatory and mixed interventions precluded a comparison of the approaches for several variables. Future research needs to describe the contents of their interventions in greater detail to establish the active ingredients in these treatments.