Does practice have an effect on neural activation? - Jolles et al. - 2010 - Article

Our working memory consists of our abilities to gather information and manipulate it for a short period of time. Research has shown that activity within the working memory corresponds with a change in activity of neurons in the frontoparietal network. Different studies found both increased and decreased activation. It is thought that initial strategies for learning to use a new strategy causes an increase in neural activity, because new neural pathways have to be created. On the other hand, when there are implemented strategies, practice with complex cognitive tasks (such as training the working memory through recalling) cause a decrease in neural activity.

Previous studies have not yet managed to get an integral picture on the effects of working memory training and report findings that contradict eachother. Task characteristics and the time window of practise both seem to have an influence on the results found. The present study has a few goals. First of all, they wanted to compare different working memory demands within a single practice paradigm. Secondly, they studied the effects of extended practices compared to a control group that was familiar, but didn’t practice the task.

The study

29 healthy adults were split up into two groups, the practice group and the control group. There were no differences in age, gender distribution and estimated intelligence between the groups. Both groups participated in two test sessions, with a time frame of six weeks between the sessions. During both sessions, fMRI data were collected while the participants performed a verbal working memory task. This task consisted of recalling three, four or five visually presented objects, that were shown and separated from each other for a fixed period of time. Participants had to verbally recall the shown objects, either forwards (maintaining the information) or backwards (manipulating the information). Within the six weeks, the practice group practised the task three days a week, for 25 minutes per practice. One of these practice session was supervised and fMRI data were collected. During the second session, participants performed the same verbal working memory task as during the first session. They also performed a set of transferring tasks which were used as indices for complex executive functioning. Six months after the second session there was a follow-up session, which included both the verbal working memory task and the transfer tasks.

Results

The percentage of correct responses and the response time on correct trials were used as a measure for working memory performance. During the first session, it was found that the accuracy of the participants decreased when they were presented with more objects. At the same time, the response time increased. There were no differences found between the practice group and the control group.

During the practice sessions it was found that the participants were more accurate and needed less time after the six weeks of practice. Performance increases were the largest within the first two weeks. The effects were the biggest for the conditions in which four or five objects were presented.

When comparing the practice group with the control group after the six week period (second session) showed that the practice group was overall more improved. This was especially the case for the conditions in which four or five objects were presented. At the follow-up, six months later, the accuracy had remained stable in the practice group. The control group showed improvement in comparison to the second session, but the practice group still performed better. There was no evidence found for transfer of skills. In other words, no differences were found between the practice group and the control group in their performance on the transfer tasks.

fMRI results

It was found that maintenance and manipulation of information in the working memory activated a frontoparietal network. There was more activation when the information had to be manipulated compared to the maintenance of the information. However, this effect disappeared when five objects were presented. This might indicate that the strategies that participants had to use to either remember or manipulate the five presented objects are more complex and therefore cause more neural activity. During the second session it was found that neural activation patterns hadn’t change in comparison with the first session for the conditions in which three or four objects were presented. Manipulation trails showed greater activity than the maintenance trails. When five objects were presented manipulation trials increased relative to activation in maintenance trails within the frontoparietal network.

Whole-brain interaction analyses showed increased activation in medial prefrontal regions for maintenance trails after practice and increased activation in the striatum for manipulation trails after practice. The increased striatal activation is considered to be associated wirh increased involvement of specific task-related processes or to processes involved in habit formation or learning. In conclusion, this indicated that there is a specificity of neural practice effects.