The Efficacy of Cognitive Training on Neuropsychological Outcomes in Mild Cognitive Impairment: A Meta-Analysis

Abstract

Mild cognitive impairment (MCI) or mild neurocognitive disorder is an intermediate stage of cognitive impairment between normal cognitive aging and dementia. Given the absence of effective pharmacological treatments for MCI, increasing numbers of studies are attempting to understand how cognitive training (CT) could benefit MCI. This meta-analysis aims to update and assess the efficacy of CT on specific neuropsychological test performance (global cognitive functioning, short-term verbal memory, long-term verbal memory, generativity, working memory, and visuospatial abilities) in individuals diagnosed with MCI, as compared to MCI control groups. After searching electronic databases for randomized controlled trials, 31 studies were found including 2496 participants. Results showed that CT significantly improved global cognitive functioning, short-term and long-term verbal memory, generativity, working memory, and visuospatial abilities. However, no significant effects were observed for shifting, abstraction ability/concept formation, processing speed, and language. The mode of CT had a moderating effect on abstraction ability/concept formation. The findings provide specific insights into the cognitive functions influenced by CT and guide the development of tailored interventions for MCI. While CT holds promise, further research is needed to address certain cognitive deficits and assess long-term effects on dementia progression.

Keywords: cognitive functions; cognitive training; meta-analysis; mild cognitive impairment; mild neurocognitive disorder.

Figure 1

Flowchart of the selection process of primary studies.

Figure 2

Forest plot for global cognitive functioning domain illustrating comparison between individuals with MCI involved in cognitive training (+CT) versus those involved in control training (−CT), displaying effect size (Hedges’ g) calculated using a random-effects model. ES = effect size; CI = confidence intervals; V = variance; N = total number of participants [30,33,35,36,40,41,42,46,52,54,56,58].

Figure 3

Forest plot for short-term verbal memory (memory domain) illustrating comparison between individuals with MCI involved in cognitive training (+CT) versus those involved in control training (−CT), displaying effect size (Hedges’ g) calculated using a random-effects model. ES = effect size; CI = confidence intervals; V = variance; N = total number of participants [29,31,33,34,40,43,50,59].

Figure 4

Forest plot for long-term verbal memory (memory domain) illustrating comparison between individuals with MCI involved in cognitive training (+CT) versus those involved in control training (−CT), displaying effect size (Hedges’ g) calculated using a random-effects model. ES = effect size; CI = confidence intervals; V = variance; N = total number of participants [30,51,52].

Figure 5

Forest plot for shifting ability (executive domain) illustrating comparison between individuals with MCI involved in cognitive training (+CT) versus those involved in control training (−CT), displaying effect size (Hedges’ g) calculated using a random-effects model. ES = effect size; CI = confidence intervals; V = variance; N = total number of participants [35,36,39,45,47,50].

Figure 6

Forest plot for abstraction ability/concept formation (executive domain) illustrating comparison between individuals with MCI involved in cognitive training (+CT) versus those involved in control training (−CT), displaying effect size (Hedges’ g) calculated using a random-effects model. ES = effect size; CI = confidence intervals; V = variance; N = total number of participants [33,34,38,49].

Figure 7

Forest plot for generativity (executive domain) illustrating comparison between individuals with MCI involved in cognitive training (+CT) versus those involved in control training (−CT), displaying effect size (Hedges’ g) calculated using a random-effects model. ES = effect size; CI = confidence intervals; V = variance; N = total number of participants [29,35,36,40,51,52,54,59].

Figure 8

Forest plot for working memory (executive domain) illustrating comparison between individuals with MCI involved in cognitive training (+CT) versus those involved in control training (−CT), displaying effect size (Hedges’ g) calculated using a random-effects model. ES = effect size; CI = confidence intervals; V = variance; N = total number of participants [31,34,36,40,43,50,59].

Figure 9

Forest plot for processing speed (executive domain) illustrating comparison between individuals with MCI involved in cognitive training (+CT) versus those involved in control training (−CT), displaying effect size (Hedges’ g) calculated using a random-effects model. ES = effect size; CI = confidence intervals; V = variance; N = total number of participants [35,36,39,47,50].

Figure 10

Forest plot for visuospatial and constructional ability illustrating comparison between individuals with MCI involved in cognitive training (+CT) versus those involved in control training (−CT), displaying effect size (Hedges’ g) calculated using a random-effects model. ES = effect size; CI = confidence intervals; V = variance; N = total number of participants [33,43,52,58].

Figure 11

Forest plot for language domain illustrating comparison between individuals with MCI involved in cognitive training (+CT) versus those involved in control training (−CT), displaying effect size (Hedges’ g) calculated using a random-effects model. ES = effect size; CI = confidence intervals; V = variance; N = total number of participants [29,35,36,38,40,46,47,49,50,51,54].