Cognitive transfer and durability of computerised cognitive training in older adults with mild cognitive impairment: a systematic review and meta-analysis of randomised controlled trials
Related Experiment Video
Updated: Oct 31, 2026
Working Memory Training for Older Participants: A Control Group Training Regimen and Initial Intellectual Functioning Assessment
Published on: September 20, 2020
Hao Luo1, Lin Deng1, Lingyun Long2
1Department of Physical Education, Chongqing Electric Power College, Chongqing, China.
Background:
Computerised cognitive training (CCT) may improve cognition in mild cognitive impairment (MCI), but whether gains transfer beyond trained processes and persist after training cessation remains uncertain. We examined immediate and sustained cognitive near transfer, cognitive far transfer, and functional far transfer.
Methods:
Seven databases were searched from inception to 1 July 2026. Randomised controlled trials comparing CCT with eligible controls and reporting transfer outcomes were included. Outcomes within 4 weeks after training were classified as immediate and those at least 12 weeks after cessation as sustained. Multiple eligible outcomes from the same trial were aggregated to one trial-level estimate. Random-effects models used restricted maximum likelihood and modified Knapp-Hartung inference. RoB 2 and GRADE assessed risk of bias and certainty.
Results:
Eleven reports from eight independent trials (722 randomised participants) were included; no trial was at overall low risk of bias. Immediate cognitive near transfer was small and imprecise (6 trials, N = 376; Hedges g = 0.14, 95% CI - 0.23 to 0.50; very low certainty). Immediate cognitive far transfer was also uncertain (8 trials, N = 586; g = 0.23, 95% CI - 0.35 to 0.81; very low certainty) and strongly influenced by Wu et al.; excluding this trial yielded g = 0.00 (95% CI - 0.26 to 0.25). Sustained cognitive near transfer (4 trials, N = 287; g = -0.07, 95% CI - 0.31 to 0.18) and sustained cognitive far transfer (5 trials, N = 397; g = 0.02, 95% CI - 0.26 to 0.29) were close to the null but based on few trials. Functional far-transfer evidence was low-certainty and limited.
Conclusion:
The evidence base is small and methodologically limited, and transfer classification requires reviewer judgement. Current evidence does not establish convincing durable transfer effects beyond trained processes. Larger, rigorous trials with validated functional outcomes and longer follow-up are needed.
Systematic Review Registration:
https://www.crd.york.ac.uk/prospero/, identifier CRD420261463612.
