Working memory : development, disorders and training

Abstract

Working memory (WM) is the ability to keep information online during a short period of time. Brain regions underlying WM functioning are found in the frontal and parietal cortices. It is largely unknown to what extent the neural substrates underlying WM are susceptible to training induced change. Here we investigate the development of WM capacity, if improvement by training is possible and explore the neuronal correlates for training induced change.

In Study I we used functional magnetic resonance imaging (fMRI) to investigate the neural correlates of the developmental change in WM capacity during childhood. We found that during performance of a visuo-spatial working memory test (VSWM), there was a significantly higher activity in the superior frontal and intraparietal cortex in subjects with higher capacity. Thus, the development of these areas may underlie the development of VSWM during childhood.

In Study II we used the VSWM test in children with and without ADHD and found that the test differentiated between these groups (P<.01). This supports the hypothesis that the WM deficit is a core deficit in ADHD. We proceeded by investigating if it was possible to improve WM capacity by training on WM tasks and secondly, if training effects would generalize to other cognitive tasks and areas of behaviour. These hypotheses were tested by designing a computerized program for WM training which was implemented in children with ADHD in two consecutive studies.

In Study III (N=14), we saw significant improvements in the treatment group as compared to the control group on the trained WM task (P < .001), on the non-trained WM tests (span-board P < .001, and digit span P < .001), as well as on other non-trained tasks; Stroop (P < .01), Raven (P < .001).

In Study IV the effects on the cognitive tests were replicated at a significance level of .01, with a randomized, double-blind, placebo-controlled multi-center design (N=44). In this study we also found generalisation of training effects to the behavioural level as evaluated by parents and teachers (inattention (P < .01) and parentrated hyperactivity/impulsivity (P < .05).

In Study V we investigated if stroke patients with significant WM deficits also could benefit from training. Participants suffering stroke one to three years earlier gained significant improvements in WM capacity (digit span p < .005, span board P < .05, and PASAT P < .00 1), and attention (RUFF 2&7 p < .005), as well as on cognitive symptoms in daily life, as measured by a self rating questionnaire (P < .005).

Study VI was undertaken to explore the neuronal correlates of WM improvement. Healthy young adults underwent fMRI before and after WM training. Task specific increases in brain activity were found in prefrontal and parietal cortices. These regions are known to underlie WM functioning.

Summary: WM shows a prolonged developmental course in humans. WM deficits are prominent in ADHD and following brain injury. However, WM can be improved by training and the treatment effect also generalizes to other cognitive tasks. The increase in WM capacity during childhood as well as after training is associated with increased brain activity in the prefrontal and parietal cortex.