Abstract
BACKGROUND: Episodic thinking, whether past or future, uses similar neural machinery, and individuals with alcohol dependence have clear challenges with both. Moreover, alcohol-dependent individuals' narrowed temporal window likely gives rise to greater valuation of immediate rewards. We aimed to strengthen working memory (WM) in alcohol-dependent individuals and measure performance on near-transfer (novel WM) and far-transfer delay discounting (DD) tasks, including episodic future thinking (EFT) performance. Importantly, heterogeneous intervention responses could obscure a treatment effect due to individuals' baseline differences. Therefore, we considered WM, DD, and EFT DD scores using rate-dependent analyses. METHODS: A total of 50 alcohol-dependent individuals received either 20 active (Trained) or sham (Control) WM training sessions using the Cogmed adaptive WM training program. Participants completed a near-transfer novel WM task and far-transfer DD and EFT DD tasks before and after training. RESULTS: Active WM training improved performance on the near-transfer task. As determined by Oldham's correlation [r(mean(x,y),)(y)(-)(x)], initially low near-transfer task scores improved more than initially high scores (i.e., rate dependence) in the Trained group only. Moreover, Trained group individuals with the highest rates of EFT DD at baseline rate-dependently decreased following training, whereas WM training had no effect on DD alone. CONCLUSIONS: These data support the notion that WM training improves near-transfer task performance and may enhance the effects of EFT DD in a subset of alcohol-dependent individuals trapped within the narrowest temporal window. Rate-dependent changes highlight that we should attend to baseline performance to better identify individuals who would most benefit from an intervention.