Which older adults maintain benefit from cognitive training? Use of signal detection methods to identify long-term treatment gains.

BACKGROUND: Cognitive training has been shown to improve memory in older adults; however, little is known about which individuals benefit from or respond best to training in the long term. Identification of responders' characteristics would help providers match cognitive interventions to individuals to improve their effectiveness. Signal detection methods may prove more informative than more commonly used analytic methods. The goal of the current study is to identify baseline characteristics of long-term treatment responders and of those able to maintain their initial benefit from cognitive training.
METHODS: Participants were 120 non-demented, community-dwelling older adults who had participated in a cognitive training intervention. Tested predictors included both demographic and neurocognitive variables. Primary outcome variables were performance on measures of memory at one-year follow-up.
RESULTS: Results of the signal detection analysis indicated that different neurocognitive performances predicted long-term effects of memory training and maintenance of initial treatment response according to different types of to-be-remembered material. Higher baseline scores on tests of associative memory, delayed verbal memory, attention, episodic memory, and younger age were found predictive of long-term response one year later. Higher associative memory scores and lower initial gains at the end of treatment (week 14) predicted successful maintenance of training gains at week 52.
CONCLUSIONS: To derive long-term benefit from particular cognitive training programs, it appears necessary for older adults to have specific neurocognitive profiles. Further, inclusion of booster sessions to cognitive training programs may assist in maintenance of initial treatment gains.