Epilepsia. 2015 Nov;56(11):1784-92. doi: 10.1111/epi.13195.

Near-transfer effects following working memory intervention (Cogmed) in children with symptomatic epilepsy: An open randomized clinical trial.

Kerr EN1,2, Blackwell MC1.
  • 1The Hospital for Sick Children, Toronto, Ontario, Canada.
  • 2University of Toronto, Toronto, Ontario, Canada.



Objective: Limited research exists regarding the effectiveness of educational and psychological interventions for improving commonly presenting cognitive impairments experienced by children with epilepsy. We evaluated the efficacy of a commercially available, computerized, working memory (WM) program using a well-defined population of children with epilepsy.

Methods: In this controlled trial, 77 children with symptomatic epilepsy (ages 6-½ to 15-½ years; 100% taking medication) with estimated intellectual ability greater than the 2nd percentile were randomly assigned to an Intervention (n=42) or Waitlist-control (n=35) group. Standardized assessments of attention and WM were administered pre- and post- training or waitlist interval, seven weeks apart.

Results: Without intervention, participants displayed significant weaknesses in intelligence, attention, and WM compared to normative samples. Controlling for pre-intervention scores and intelligence, significant treatment effects for the Intervention group were found for visual attention span, auditory WM, and visual-verbal WM. Intention-To-Treat analyses (all participants) and sensitivity analyses (n=37 and n=21 for the Intervention and Waitlist-control groups, respectively) were highly similar, providing confidence to the results. Effect sizes for significant outcomes were large  (> two-thirds of the standard deviation of the normative-data).  The clinical/demographic and functional factors studied did not elucidate who most benefits from training.

Conclusion: This is the first study to evaluate the effectiveness of intervention to ameliorate WM deficits commonly experienced by children with symptomatic epilepsy. Results support group improvement on some untrained tasks immediately post-intervention, demonstrating preliminary usefulness of Cogmed as a treatment option.

KEYWORDS: Attention; Children; Epilepsy; Intervention; Working memory

PMID: 26391621



Epilepsy is a common neurological condition (1 in 100) characterized by at least two unprovoked seizures (i.e., transient symptoms associated with abnormal excessive or synchronous neuronal activity in the brain) or one unprovoked seizure with a heightened likelihood of additional seizures. However, Epilepsy is far more than a seizure disorder, particularly for those who have ongoing seizures despite medication. Apart from the recurrent seizures, children are at increased risk for cognitive, behavioural, and psychosocial difficulties compared to peers in the general population, and even to those with other chronic health conditions. Risk factors include effects of chronic seizures, ictal and post-ictal effects (i.e., effects during and after the seizure), effects of interictal discharges (between seizures), morphological factors (e.g., altered brain networks), clinical factors (e.g., early age of seizure onset, more than one seizure a day, duration of seizures), and possible side effects of anti-epileptic drugs.

Attention-Deficit/Hyperactivity Disorder (ADHD) is one of the most frequently observed co-occurring diagnosed conditions in those with Epilepsy (see Figure 1 for differences in Primary ADHD and ADHD presentation in Epilepsy). In addition, difficulties with attention, particularly sustained attention, as well as working memory are among the most frequently reported cognitive processing issues in both conditions. Working memory refers to the ability to mentally retain, manipulate and integrate bits of information for a brief period of time. Working memory may be likened to a mental scratchpad or workplace that provides a temporary holding store for relevant information which is then easily available for processing and goal-directed behavior. Working memory is required for higher-order cognitive tasks including reasoning, learning, problem solving, and organizing information in our minds. It is also related to many activities of daily living and academic success. Conceivably, rehabilitative efforts aimed at improving working memory could have far-reaching benefits for children with Epilepsy. To date, however, there is very little research examining the effectiveness of educational or psychological (i.e., non-pharmacological) interventions for this patient population.



Figure 1 Symptom presentation of ADHD


A review of working memory interventions showed that Cogmed, a research-based and commercially-available working memory computer program, was most effective in improving working memory as measured on standardized tests. Additionally, Cogmed has been found to be associated with changes in neural networks within the brain that support working memory skills in typically developing children and in adults. Cogmed is a targeted intervention designed to rigorously tax a person’s working memory capacity. Throughout the course of the program (typically 5 days per week for 5 weeks), the individual engages in self-adjusting “games” to challenge him or her on a trial-by-trial basis. In addition to the working memory exercises, the program has built-in features to help the individual stay motivated, as well as requires supervision by a training aide (i.e., parent) and weekly coaching by a trained Cogmed coach. As such, the program utilizes key components for effective change; namely, targeting an area of weakness, as well as frequent and repetitive practice. The efficacy (i.e., did it work in the “ideal” or laboratory setting) and effectiveness (i.e., did it work through practitioners and at home) with ADHD is promising, but the usefulness for individuals with Epilepsy, an ongoing neurological condition with separate underlying biological markers, had not previously been evaluated.

We recruited 77 children with ongoing seizures despite medical intervention to participate. Through a coin toss, they were either placed on a waiting list and assessed seven weeks apart with no intervention, or placed in the intervention group in which they received intervention and were assessed pre and post intervention, approximately 7 weeks apart.

As per the standards for randomized control studies, intention-to-treat was our primary analysis given that there were some individuals who dropped out of the study (15 for the waitlist group and 4 for the intervention group). Intention-to-treat means that data for each individual who dropped out of the study were generated. In order to reduce the bias of individuals who dropped out and assume that their results would be exactly the same, a computer program randomly created the missing data based on the distribution of scores from the remaining participants.

We focused on the near-transfer effects of Cogmed intervention, meaning that we were interested in looking at standardized measures that assessed the same constructs of attention and working memory but were untrained (i.e., differed from the Cogmed intervention exercises in content and task format). We observed significant near transfer effects, with the intervention group showing significant improvements on three of nine measures, whereas there was no change for the waitlist group. We did not investigate the broader scope of transferrable effects, such as reasoning skills or academic achievement. In secondary analyses, we only examined the information we had on the individuals who remained in the study. Our results yielded the same conclusions. Intervention improved visual attention span (i.e., the number of chunks of visual information that could be immediately retained and processed), as well as auditory working memory and visual-verbal working memory. The measured post-intervention changes were large and represented clinically meaningful improvements. Because the waitlist group did not display any changes, the observed improvements do not appear to be related to the passage of time, the use of unreliable tests, or the benefit of practice on our tests, but rather suggest a positive effect of intervention.

At an individual level, a large proportion of individuals made significant improvements on numerous variables following the intervention. There was a wide range of improvement on trained and untrained tasks, which suggests a differential response to treatment at the individual level. Unfortunately, models used to predict those improvements did not yield meaningful results. In contrast to our predictions, differences in seizure type, age of seizure onset, duration of seizures, frequency of seizures, and type and number of medications did not have modulating effects. Despite having a relatively large sample, heterogeneity in these factors limited our understanding of their modulating influence on results.



Figure 2 Frequency of Significant Change for Intervention Participants


This table represents the proportion of individuals in the intervention group who displayed significant improvement relative to the waitlist group (i.e., the group who did not receive intervention but were assessed over the same time frame). As a group, the waitlist controls did not display any changes, and only or no waitlisted child showed any meaningful improvement without intervention.

Within our sample of children with epilepsy, 89% of the Intervention group successfully met our compliance criteria of completing 25 sessions within seven weeks. The ratio is similar to previous experimental trials with children with ADHD or learning needs. Nonetheless, attrition was a salient issue for 11% of our treatment group and may be related to the repetitive or challenging nature of the tasks along with logistical challenges (e.g., scheduling).

Epilepsy has broad and ongoing influences on the developing brain and on educational, occupational, and interpersonal functioning. Hence, it would be important to consider the transferrable and long-term cognitive and neurobiological effects as a result of intervention. Studies of Cogmed’s transferrable effects with other populations are mixed. Future research involving individuals with Epilepsy should include follow-up cognitive assessment to track long-term changes over time. Given our findings of immediate training benefits, in combination with extensive neuroplasticity literature and evidence of post-Cogmed-training neurobiological changes with typically developing children, future investigations of training-induced effects at the neurobiological level in children with epilepsy using functional neuroimaging is also certainly warranted.

This study is important because it is the first to assess whether the working memory deficits commonly experienced by children with symptomatic epilepsy can be improved following intervention. Although the results are preliminary, the group results support that Cogmed intervention is useful in improving performance on untrained, but related, tasks immediately after intervention, and that the benefits are displayed for a significant proportion of those who engage in the intervention.


Essential Reading:

Astle DE, Barnes JJ, Baker  K, Colclogh GL, Wollrich MW. Cognitive training enhances intrinsic brain connectivity in childhood. J Neurosci 2005:35 (16):6277-6283.

Baddeley AD. Working memory WM. Science 1992;255:556-559.

Elger CE, Helmstaedter C, Kurthen M. Chronic epilepsy and cognition. Lancet Neurol 2004;3:663–672.

Fastenau PT, Shen J, Dunn DW et al. Neuropsychological predictors of academic underachievement in pediatric epilepsy: moderating roles of demographic, seizure and psychosocial variables. Epilepsia 2004;45:1261-1272.

Holmes J, Gathercole SE, Place M, et al. Working memory deficits can be overcome: impacts of training and medication on working memory in children with ADHD. Appl Cogn Psychol 2010;24:827–836.

Holmes J, Gathercole SE. Taking working memory training from the laboratory into schools. Educ Psychol 2013;34(4):440-450.

Jonides J, Lacey SC, Nee DE. Processes of working memory in mind and brain. Curr Dir Psychol Sci 2005;14:2-5.

MacAllister WS, Vasserman M, Vekaria P, et al. Neuropsychological endophenotypes in ADHD with and without epilepsy. Appl Neuropsychol Child 2012;1:121–128.

Melby-Lervag M, Hulme C. Is working memory training effective? a meta-analytic review. Dev Psychol 2013;49:270–291.

Reilly C, Atkinson P, Das KB, et al. Neurobehavioral comorbidities in children with active epilepsy: a population-based study. Pediatrics 2014;133:e1586-93.

Sherman EM, Slick DJ, Connolly MB, et al. ADHD, neurological correlates and health-related quality of life in severe pediatric epilepsy. Epilepsia 2007;48:1083–1091.

Shinaver CS, Entwistle PC, Söderqvist S. Cogmed WM training: reviewing the reviews. Appl Neuropsychol Child 2014;3:163-72.



Multiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier SchönmannMultiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier Schönmann