Memory focused cognitive training for older people with MCI: A CogTale Meta‐Analysis

Background Cognition‐oriented treatments (COTs) such as cognitive training (CT) are effective for improving cognition in people with MCI. Although multi‐domain CT has positive effects on global cognition, the extent to which memory‐focused CT impacts memory functions more specifically, and how process‐based, strategy‐based, and group‐based CT studies compare remains unknown. Method We conducted a systematic review using CogTale (Sabates et al. 2021), a novel evidence synthesis tool which automates aspects of the meta‐analysis pipeline. Embase, Medline and PsycINFO databases were searched from inception to 19/10/2022. Studies reporting the results of a controlled, memory‐oriented CT intervention for people with MCI, and that reported at least one cognitive outcome were eligible. Multi‐domain CT, multicomponent interventions were excluded. Data from eligible studies was extracted into the CogTale platform, and methodological quality scores were computed for each study. Random effects inverse variance meta‐analysis was performed for all outcomes reported by a minimum of three studies, and confidence in the effect estimates was computed based on heterogeneity, precision, and risk of bias. Result After removing duplicates, 653 studies were screened and 101 studies underwent full‐text review, of which 24 (17 reporting any strategy, 16 individually‐delivered) were eligible. Across studies, small to moderate effects were found for auditory‐verbal, visuospatial, and mixed memory outcomes, and for working memory. Findings were similar when only studies using any memory‐strategy were included (k = 17). However, when studies using only mnemonic strategy training (MST) were analysed (k = 8), effect estimates for the small number of outcomes that could be pooled did not reach significance, but the confidence in these findings was low. Conclusion Memory‐oriented CT studies are associated with gains in measures of memory and working memory but not on measures of global cognition in people with MCI. Although the analysis suggested that the use of any‐memory strategy (e.g., MST, errorless learning, spaced retrieval) were associated with similar effects, when examined in isolation, no evidence was found that MST‐based CT is associated with gains on any outcomes studied. More high‐quality studies of MST‐based CT are needed to determine whether this effortful approach to CT is beneficial relative to less effortful approaches such as rehearsal‐based CT.