Open Access

Mandy Roheger, Elke Kalbe, Anne Corbett, Helen Brooker, Clive Ballard

• Abstract Background Identifying predictors for general cognitive training (GCT) success in healthy older adults has many potential uses, including aiding intervention and improving individual dementia risk prediction, which are of high importance in health care. However, the factors that predict training improvements and the temporal course of predictors (eg, do the same prognostic factors predict training success after a short training period, such as 6 weeks, as well as after a longer training period, such as 6 months?) are largely unknown. Methods Data (N = 4,184 healthy older individuals) from two arms (GCT vs. control) of a three‐arm randomized controlled trial were reanalyzed to investigate predictors of GCT success in five cognitive tasks (grammatical reasoning, spatial working memory, digit vigilance, paired association learning, and verbal learning) at three time points (after 6 weeks, 3 months, and 6 months of training). Possible investigated predictors were sociodemographic variables, depressive symptoms, number of training sessions, cognitive baseline values, and all interaction terms (group*predictor). Results Being female was predictive for improvement in grammatical reasoning at 6 weeks in the GCT group, and lower cognitive baseline scores were predictive for improvement in spatial working memory and verbal learning at 6 months. Conclusion Our data indicate that predictors seem to change over time; remarkably, lower baseline performance at study entry is only a significant predictor at 6 months training. Possible reasons for these results are discussed in relation to the compensation hypothesis. J Am Geriatr Soc 68:‐, 2020.