Locating neural transfer effects of n-back training on the central executive: a longitudinal fMRI study

The large number of behavioral studies testing whether working memory training improves performance on an untrained task have yielded inconclusive results. Moreover, some studies have investigated the possible neural changes during the performance of untrained tasks after training. Here, we studied the transfer from n-back training to the Paced Auditory Serial Addition Test (PASAT), two different tasks that use the central executive system to maintain verbal stimuli. Participants completed fMRI sessions at baseline, immediately after one week of training, and at the five-week follow-up. Although behavioral transfer effects were not obtained, training was associated with decreased activation in the anterior dorsolateral prefrontal cortex (DLPFC; BA 9/46) while performing the PASAT that remained stable five weeks later. Consistent with our hypothesis, the changes in the anterior DLFPC largely overlapped with the n-back task fMRI activations. In conclusion, working memory training improves efficiency in brain areas involved in the trained task that may affect untrained tasks, specifically in brain areas responsible for the same cognitive processes.

Participants watched the letters, instructions, and fixation point (all black ink with a 54point Arial font )in the middle of the screen on a white background.
To familiarize themselves with the stimuli presentation and how to respond, they performed a five-minute practice task outside the scanner that was composed of three blocks, one per load level. A similar laptop with the same display configuration and the hardware as the one used to present in-scanner task was used for the oral responses.
Subjects were told to answer as quickly as possible, but avoiding making mistakes, and they were given oral instructions about how to perform the task.

N-back training task
Four consecutive training sessions (TS) of single n-back were conducted by the trained group after fMRI in S1. They came to our laboratory located at the University and carried out only one TS per day. Each TS were organized in two phases and lasted 60 minutes: the learning part and the testing part. In the learning part that lasted 50 minutes, subjects carried out an adaptive n-back paradigm adapted from (Jaeggi, Buschkuehl, Jonides & Perrig, 2008), whereas in the testing part, they performed a simple n-back task that lasted 10 minutes. Hence, the total training time was approximately 200 minutes, plus 40 minutes for the testing part. Their results on the test part were useful to evaluate their progress on the task. The same laptop as in the fMRI sessions was used, and their responses were collected via response-grips (NordicNeuroLab, Bergen, Norway).
The adaptive n-back task was composed by three active load levels (1-, 2-and 3-back) using the same stimuli and block timing as in the n-back fMRI task and lasted approximately 16 minutes. Participants carried out three runs per TS. Feedback were given to the subjects after each stimulus and at the end of each block about their accuracy and time reaction. For motivational reasons, we changed the level of difficulty by changing the level of "n" (1, 2, or 3). After each block, the participant's individual performance was analyzed, and the n-back level was automatically adjusted up to a maximum of 3-back. Thus, if the participant had at least 90% correct answers, the level of "n" in the next block was increased by one, but it was decreased by one if accuracy was below 80%. The n-level remained constant in all other cases. In the last run, we increased the percentage by five percent to make it more difficult. Therefore, if the participant had at least 95% correct answers, the level of "n" was increased by one, whereas it was decreased by one if accuracy was below 85%. Each run started with the minimum level of "n". For the feedback: a colored circle appeared for a few seconds at the corner of the screen after each response: green if the answer was correct, red if it was a mistake, and blue in omissions. Also, correct response percentage and reaction time average was given to participants at the end of each block. On the test part which lasted 10 minutes and without feedback, subjects carried out an eight-block n-back task, half of 2-back and half of the 3-back. To each stimulus, the E-Prime software collected each participant's accuracy and reaction time (RT).

N-back fMRI behavioral analysis
In order to process the behavioral data (accuracy and RTs for participants' performance)

RESULTS
These data have been reported in our previous study (40).

N-back behavioral fMRI results
The repeated measures 2x3x3 mixed-model ANOVA conducted for accuracy yielded main effects for Session (F(2,50) = 34.66 p<.001) and Load Level (F(2,50) = 42.85 p<.001), which means that all the participants reduced their mistakes in the post-training and follow-up sessions, compared to Session 1, and that the highest accuracy scores were observed during the 0-back. These main effects were driven by significant Group x Session (F(2,50) = 7.77 p=.001), Load Level x Session (F(4,48) = 13.07 p<.001), and Load Level x Group (F(2,50) = 7.23 p=.002) interactions. The first interaction indicated that trained participants were better than controls during the post-training and follow-up sessions. The second indicated that differences between load levels were greater at pretraining, whereas the third reflected that the training group showed better performance than the control group on 2-back and 3-back. As expected, the Load Level x Session x Group interaction reached significance (F(4,48) = 4.01 p=.007), which means that the trained group became more accurate in the post-training and follow-up sessions than the control group, when performing the 2-back and 3-back load levels (see Supplementary   Fig. S1). Post-hoc analyses revealed that these differences were significant for 3-back vs 0-back (p=.002), and they only approached significance (p=.13) for 2-back vs 0-back (40).
Analyses of RTs scores revealed a similar pattern to the one found for accuracy. The 2x3x3 ANOVA also yielded significant main effects for Session (F(2,50) = 51.59 p<.001) and Load Level (F(2,50) = 75.37 p<.001). Both groups responded faster in the post-training and follow-up sessions than in the pre-training session. Moreover, both responded faster in the 0-back load level compared to the 2-back load level, as well as in the 2-back load level compared to the 3-back load level. Significant two-way interactions were obtained for the Group x Session (F(2,50) = 28.14 p<.001) and Load Level x Session (F(4,48) = 28.23 p<.001) interactions. The first two interactions may be interpreted similarly to accuracy; participants were faster than controls during the post-training and follow-up, and the differences between load levels were greater at pre-training. Importantly, all these significant effects were qualified by the three-way Load Level x Session x Group interaction, which was highly significant (F(4,48) = 11.34 p<.001). As expected, this interaction showed that the training group, compared to the controls, was faster after training and in the follow-up session in the 2-back and 3-back load levels (see