Cognitive training for Alzheimer's disease and other forms of dementia: Insights from a systematic review and Bayesian meta-analysis

Background The prevalence of individuals living with mild cognitive impairment (MCI), Alzheimer's disease (AD), and other forms of dementia is globally increasing. Four out of nine international clinical guidelines recommend non-pharmacological cognitive interventions to enhance cognition, independence, and wellbeing. However, the effectiveness of cognitive rehabilitation (CR) and cognitive training (CT) for individuals with MCI and AD and other forms of dementia is still debatable, often due to significant heterogeneity among studies. Objective This study aims to assess the effectiveness of CT and CR in these populations. Methods Following PRISMA guidelines, we conducted a comprehensive literature search across databases including OVID, MEDLINE, EMBASE, and Scopus, identifying randomized controlled trials and non-randomized pre-post intervention studies. The Bayesian meta-analysis focused on pre-post changes in global cognition, quality of life, everyday functioning, and depression, avoiding comparisons with control groups to reduce heterogeneity (PROSPERO: CRD42022365038). Results The search yielded 6075 results, with 40 studies meeting inclusion criteria, encompassing 50 independent trials. CT and people with AD and other dementias are the best represented intervention and population, respectively. CT was more effective in improving global cognition in individuals with AD and other dementias, and paper-and-pencil and face-to-face formats yielded greater benefits. The analysis showed a significant susceptibility to bias among the studies. Conclusions Limitations in outcome measure (e.g., MMSE) suggest the need for more sensitive assessments, especially for MCI. Future research should explore broader aspects of wellbeing and investigate the potential of CR. Policymakers are encouraged to consider these findings when designing cognitive interventions for this population.