Predictors of Improvement after Cognitive Training in Mild Cognitive Impairment: Insights from the Cognitive Training and Neuroplasticity in Mild Cognitive Impairment Trial

Objective: Cognitive training may benefit older adults with mild cognitive impairment (MCI), but the prognostic factors are not well-established. Methods: This study analyzed data from a 78-week trial with 107 participants with MCI, comparing computerized cognitive training (CCT) and computerized crossword puzzle training (CPT). Outcomes were changes in cognitive and functional measures from baseline. Linear mixed-effect models were used to identify prognostic factors for each intervention. Results: Baseline neuropsychological composite z-score was positively associated with cognitive and functional improvements for both interventions in univariable models, retaining significance in the final multivariable model for functional outcome in CPT ( P < 0.001). Apolipoprotein E e4 carriers had worse cognitive ( P = 0.023) and functional ( P = 0.001) outcomes than noncarriers for CPT but not CCT. African Americans showed greater functional improvements than non-African Americans in both CPT ( P = 0.001) and CCT ( P = 0.010). Better baseline odor identification was correlated with cognitive improvements in CPT ( P = 0.006) and functional improvements in CCT ( P < 0.001). Conclusion: Baseline cognitive test performance, African American background, and odor identification ability are potential prognostic factors for improved outcomes with cognitive interventions in older adults with MCI. Apolipoprotein E e4 is associated with poor outcomes. Replication of these findings may improve the selection of cognitive interventions for individuals with MCI.