The effect of working memory training on patient and informant reported executive function in mild cognitive impairment: an interventional study.

Abstract

Background: Self-reported outcome measures are rarely described in individuals with mild cognitive impairment (MCI). In this study, we investigated the effect of computerized working memory training on self- and relatives reported executive function measures.

Methods: A total of 50 individuals with MCI were recruited from five memory clinics in Southern Norway and underwent a 5-week/20-25 sessions computerized working memory training program. Both individuals and relatives scored the "Behavior Rating Inventory of Executive Function for Adults" (BRIEF-A) before, and at least four months after finalizing training. Statistical analyses included paired t-tests and mixed ANOVA with rater type (participant/informant) and time (baseline/4-months).

Results: Mixed ANOVA revealed significant (adjusted p ≤ 0.025) participant-informant x time interactions for Working Memory, Metacognition Index and Global Executive Composite, suggesting that the changes in ratings over time differed between participants and informants. Post hoc tests revealed that at baseline participants rated themselves as significantly more impaired (adjusted p ≤ 0.036) than informants on the Working Memory scale, the Metacognitive Index and the Global Executive Composite. After training, no significant differences were found between participant and informant reports. After training participants reported significant improvements on Working Memory (adjusted p = 0.038), Metacognition Index (adjusted p = 0.038), with working memory improving from above the established clinically significant impairment threshold to the normal range. The change in Global Executive Composite was not significant after correction for multiple comparisons (adjusted p = 0.057). No significant changes were found for relatives' reported scores between the two time points. Excluding participants developing dementia during the study enhanced the difference in mean values before and after the intervention for these outcomes.

Conclusion: Computerized working memory training significantly improved self-reported executive function in individuals with MCI. While informant reports remained stable, the prior significant discrepancy between self- and informant ratings converged after training due to this improvement. These findings suggest that training can enhance subjective cognitive experience and awareness, offering potential clinical benefits for individuals with MCI. Our study highlights self-report measures as valuable and sensitive outcomes in MCI interventions, capturing the personal experiences of cognitive change.

Trial registration: ClinicalTrials.gov, NCT01991405. Registered on 18.11.2013.