The main aim of this prospective interventional study is to compare quantitative EEG findings and cognitive and neuropsychological test results before and after training with the Cogmed JM program in a group of very-low-birth-weight children in preschool age, i.e. ages 5-6 years. The investigators hypothesize that the children in the study may respond positively to the computerized training and improve working memory, but probably to different degrees depending on the underlying neurological condition. The investigators also hypothesize that training may benefit additional executive functions.

Children born preterm are at increased risk of neurological disabilities due to perinatal brain damage. During the last ten years an increasing number of especially extremely low birth weight (ELBW: BW ≤ 1000g) children has survived, and one might wonder whether this has lead to increased amount of brain pathology and impairments among survivors. Major impairments including cerebral palsy, mental retardation and impaired vision and hearing are most common in children born extremely preterm (gestational age below 28 weeks). In addition to this, a large part of survivors after preterm birth will experience motor problems, cognitive dysfunctions, psychiatric and behavioural problems of varying severity.

In our previous research we have found significant deficits in working memory in very preterm born children, and this seems to have a strong influence on cognitive functioning. During the last years, several studies have shown that working memory skills can be trained, and training working memory to improve cognition and executive functions is regarded as one of the major steps forward in neuroscience in recent years.

The working memory training program version for preschoolers, Cogmed JM, has recently been launched. Healthy preschoolers trained on working memory improved significantly on trained tasks but also on non-trained tests of spatial and verbal working memory, as well as transfer effects on attention. The preschool version of the program has not yet been administered to preterm born or other neurologic high risk children in this age group.

The main aim of this prospective interventional study is to compare quantitative EEG findings and cognitive and neuropsychological test results before and after training with the Cogmed JM program in a group of very-low-birth-weight children in preschool age, i.e. ages 5-6 years. We hypothesise that the children in the study may respond positively to the computerized training and improve working memory, but probably to different degrees depending on the underlying neurological condition. We also hypothesise that training may benefit additional executive functions.

The children in the study will be divided into two subgroups (15/15 children). Subgroup A will start training, while subgroup B waits during this first training period. Subgroup B will therefore act as a control group. After 8 weeks subgroup B will then start training. This is in agreement with the so-called Stepped Wedge design (Brown and Lilford 2006).

Inclusion Criteria:

• VLBW (birth weight ≤ 1500 grams)
• preschool child
• born at St. Olav's University Hospital in Trondheim in 2005 and 2006

Exclusion Criteria:

• birth weight >1500 grams
• diagnosed genetic syndromes
• severe cerebral palsy with totally impaired bilateral hand function (GMFCS - gross motor function classification scale level V)
• blindness