Computerized memory training leads to sustained improvement in visuospatial short-term memory skills in children with Down Syndrome

Publication: American Journal on Intellectual and Developmental Disabilities

Institution: University of Portsmouth, Down Syndrome Education International,MRC Cognition & Brain Sciences Unit (Cambridge)

Investigator(s): Stephanie J. Bennett, Ph.D., Joni Holmes, Ph.D., Sue Buckley, Ph.D

Program: Cogmed JM

Background & Aim: Down syndrome (DS) affects 1 in 690 live births. It is estimated that 83,400 children and adolescents in the United States were living with DS in 2002. DS is associated with significant delays in verbal short-term memory (STM), language, and speech. Differential performance on tasks designed to access the verbal and visuo-spatial storage components of working memory (WM) also provide evidence that there is a more severe deficit in verbal STM than visuo-spatial STM in DS individuals. This storage aspect of WM is vital for language acquisition and grammar development and is correlated with reading ability in typically developing persons. Although verbal STM deficits are commonly reported, individuals with DS also have impairments in the visuo-spatial domain relative to healthy age-matched controls. Establishing interventions aimed at reducing these cognitive deficits is needed so to provide evidence-based education and therapy for individuals with DS.

The aim of this research was to investigate whether Cogmed Working Memory Training could be efficacious in reducing memory difficulties associated with DS. A pilot study was conducted using Cogmed RM, the school –aged version of the program which involves training on 8 verbal and visuo-spatial WM exercises for 30 to 40 minutes per day, 5 days per week for 5 weeks. For the five children with DS, ages 7 to 10 years, that trained with Cogmed RM it was found that the school-aged program was too challenging and required 50 minutes at baseline. Thus, the current research employed the preschooler version Cogmed JM, which includes 3 visuo-spatial WM tasks per day, 5 days per week for 5 weeks. Cogmed JM has previously shown to improve both visuo-spatial and verbal WM skills in typically developing children (Thorell et al., 2009). Children in the current research were tested using the Automated Working Memory Assessment (AWMA) and parents rated behavior using the Brief Rating of Executive Function- Preschool Version (BRIEF-P).

Population & Sample Size: N = 25 children with DS, ages 7 – 12 years

• n = 10 children in adaptive Cogmed JM training group (group 1)
• n = 8 children in waitlist control group (adaptive Cogmed JM training)(group 2)
• n = 4 children were excluded due to illness, behavioral difficulties, technical difficulties, or failure to complete training
• n = 3 children lost at follow up

* Based on baseline IQ testing (Kaufman Brief Intelligence Test (KBIT)), children had a mean mental age between 4 and 7 years old.

Design: Randomized, waitlist controlled, test-retest, 4 month follow up

T1 = baseline, T2 = post-test (group 1), T3 = post-test (group 2), 4 month follow up (group 1)


I. All children improved significantly on the trained WM tasks.

II. Children in group 1:
1) Improved significantly on visuo-spatial STM and WM (AWMA) at T2. Gains maintained at follow up (T3).
2) Improved significantly from compared to the waitlist control (group 2) on visuo-spatial STM (T2).
3) Had no significant improvement on verbal STM or WM (AWMA) at T2 or T3.
4) Improved significantly on parent rated shift and WM abilities (BRIEF-P) (T2). Gains maintained at T3.

III. Children in group 2:
1) Improved on visuo-spatial and verbal STM at post-test (T3).
2) No significant improvement on verbal or visuospatial WM (T3).
3) No significant improvement on BRIEF-P subscales.

Summary and Implications: This study was the first to evaluate the feasibility and efficacy of Cogmed JM for children with DS. On average, children completed 92% of the total number of tasks presented over the duration of training, which took place at school. The main findings were that children with DS improved significantly on trained and non-trained visuo-spatial STM tasks compared to a group that was exposed to no training, with gains maintained at 4 month follow up. A partial replication of training effect was observed when the waitlist control group had comparable gains in visuo-spatial STM after training.

Contrary to expectation, children who trained did not demonstrate generalization from training on visuo-spatial tasks to verbal outcome measures. The domain specific improvements could be a consequence of verbal processing difficulties for DS children that limited the transfer of positive effects across domains in this particular population. It is also possible that the benefits of training were limited by the sample’s significant level of cognitive delay (average IQ below 70). Although some small improvements were seen in verbal STM for the waitlist control at post-test and for group 1 at 4 month follow up, it is possible these gains represent practice effect or developmental improvement. It is also possible that these gains represent emerging transfer to the verbal domain and thus, this effect warrants further investigation in future studies with long term follow up.

Some of the children improved on parent ratings of behavior as measured by the BRIEF-P. The first group to train (group 1) improved on the shift and WM scales with gains maintained at 4 month follow up. These finding were not replicated for the waitlist control group. It is possible that the improved ability to shift attention between tasks and parts of tasks for these children post training suggests that this function may be malleable in this group. It is however acknowledged by the researchers that the BRIEF-P is a subjective measure and group allocation was not blinded.

Despite some limitations of this study including a non-active control group, subjective behavioral measure, and small sample size, it is important to recognize that Cogmed was demonstrated to be feasible for children with DS and low IQ in a school setting. Unlike previous research with typical preschoolers that took place under tightly controlled experimental conditions, this study utilized school staff for support of participant training. Observations by school staff included increased cooperation, less frustration with incorrect answers, and the ability to concentrate longer in class.

Overall, training was associated with improvement in visuo-spatial STM and WM, with effects of 0.59 and 0.83 respectively. Problem behaviors were also reduced for some of the sample, particularly in the area of shifting attention. Future research should investigate the impact of WM training using objective executive function outcome measures. The findings of the research thus imply that children with DS may benefit from WM training.

Funding: This research was supported by a project grant from the Baily Thomas Charitable Trust and by Down Syndrome Education International.