Working Memory Training Improves Cognitive Function in VLBW Preschoolers

Background And Objective: Preterm born children perform poorer than term peers on tests of attention and executive functions including working memory tests. Our aim was to evaluate if preterm born preschoolers with very low birth weight (VLBW) would benefit from a computerized working memory training program and if the training would have a generalizing effect on memory, learning, attention, behavior, and anxiety.

Methods: A prospective intervention study with a stepped wedge design where 20 VLBW preschoolers aged 5 to 6 years participated. The children trained with the Cogmed JM program for 10 to 15 minutes a day, 5 days a week over a 5-week period. Extensive neuropsychological assessment and parental questionnaires regarding behavior and anxiety were performed before and 4 weeks after intervention.

Results: The children improved significantly on trained (mean Start Index 42.1 [SD 6.3]), mean Max Index 60.6 [SD 5.7]), and nontrained working memory tasks (Spatial Span backward; 2.3 [before] to 3.6 [after training] [confidence interval {CI} -2.2 to -0.4] and Spatial Span total score; 6.4-8.3 [CI -3.7 to -0.1]). A generalization effect was found on auditory attention (49.6-58.2 [CI -15.5 to -1.6]), phonological awareness (9.3-12.6 [CI -5.2 to -1.4]), visual (memory for faces 20.0-24.9 [CI -7.4 to -2.5]), as well as verbal memory (narrative memory; 12.9-17.5 [CI -7.1 to -2.0], and sentence repetition 15.7-17.7 [CI -3.3 to -0.7]).

Conclusion: This study shows that VLBW preschoolers benefit from a computerized working memory training program. We speculate that such training before starting school may prevent or reduce cognitive problems that impact educational achievement.