1229 Computerized Working Memory Training is Effective in Preterm Born Children at Preschool Age

Background and Aims Working memory (WM) is defined as the skill to retain and manipulate information “on-line” over short periods of time. Deficits in WM are frequently reported in preterm children and are associated with cognitive, behavioural and academic problems. We wanted to evaluate if a software-based computer program would improve WM, attention and behaviour in VLBW children at preschool age. Methods This prospective study included 20 preterm children. Mean birth weight 1099g(SD 311), mean gestational age 29 weeks(SD 2.8). At age 5–6, the children trained with the Cogmed JM computer program for 10–15 minutes each day, 5 days a week for 5 weeks. The children were assessed before and 4 weeks after training. Effect on trained WM tasks were assessed by improvement measures included in the computer program. Non-trained WM and generalization effects were assessed by neuropsychological (NEPSY) tests. Parental questionnaires regarding ADHD symptoms were performed before and after completed training. Results The children improved on trained WM tasks (Start-Index: mean 42.1; SD 6.3, Max-Index 60.6; SD 5.7. p≤0.001.). The group also improved on non-trained WM tasks and showed a generalization effect on auditory attention, phonological awareness and visual and verbal memory. A Stepped-Wedge-Design showed that improvements in test results after training were not due to test-retest effects. There was a trend towards significantly reduced ADHD scores after training. Conclusion Computerized WM training in VLBW pre-schoolers have positive effects on trained and non-trained WM tasks as well as generalizing effect on verbal and visual learning and memory functions.