LAUSR

BACKGROUND Improving children's sleep could lead to significant benefits in several functional domains. Recent research in adults suggests that intensive, adaptive cognitive training may be beneficial in improving sleep, although there is limited understanding whether this approach yields similar results in children. OBJECTIVE To determine whether a working memory training program improved sleep latency, sleep problems, and sleep duration on school and nonschool nights and whether there was a differential effect of the timing of training during the school day on sleep outcomes. DESIGN/METHODS Design: Population-based randomised controlled trial. SETTING Forty-four schools in Melbourne, Australia. PARTICIPANTS All Grade 1 children (mean age = 6.9 years, SD 0.4) underwent WM screening using two subtests from the Automated Working Memory Assessment. Children with low verbal and/or visuo-spatial WM scores relative to their peers ('low WM', ∼25%) were randomised to intervention or control arms. INTERVENTION 20 to 25 computerised 25-min training sessions were conducted using the CogMed program, over 5-7 weeks at school. OUTCOMES Parent-reported child sleep characteristics (time, latency, duration and problem) at 6 months post randomisation. RESULTS A total of 452 (26.0%) of 1723 children screened (64.1% of approached) met trial eligibility criteria, with 226 in each study arm. Of intervention children, 91% completed the minimum 20 days of training. Retention was 90.5% at 6 months. Adjusted regressions showed that intervention children did not have better sleep latency, duration, bedtime consistency or less sleep problems. CONCLUSION It does not appear that adaptive working memory training during the school day can be used as a novel approach to improve children's sleep attributes up to 6 months post-randomisation, regardless of the time of day training is delivered.