Brief Review of Research on Cogmed Working Memory Training

Working memory is the cognitive system that provides temporary storage and manipulation of information necessary for a variety of complex cognitive tasks (Baddeley & Hitch, 1974, 1994; Baddeley 2000, 2001). There exists an extensive body of literature exploring the relationship between working memory and IQ, attention, problem solving, planning, reading and developmental disorders in behavior, language, and math (Engle et al., 1999; Hulme & Roodenrys, 1995; Barkley, 1997; Klingberg, 2000; Gathercole & Pickering, 2000; Swanson & Sachs-Lee, 2001; Alloway & Gathercole, 2006). The underlying neurophysiological mechanisms of working memory and its association with the frontal lobes of the brain have also been well explored (Fuster, 1989, Williams & Goldman-Rakic; 1995; D’Esposito et al., 2000; Stuss & Alexander, 2000; Wager & Smith; 2003). However, despite the abundance of literature on the function of working memory, there is a dearth of knowledge concerning working memory training and its impact.

The scarcity of published literature on working memory training can be attributed to the long-held view that working memory capacity is fixed. Recently however, Dr. Torkel Klingberg and his colleagues at Karolinska Institute have challenged the historically held view of fixed capacity with their finding that working memory capacity can be increased with intensive training (Klingberg et al., 2002, 2005). Stemming from the success of their initial training studies, these researchers developed a training program that became the basis for what is now called Cogmed Working Memory Training. Currently, Cogmed is being used by researchers worldwide in order to explore how training and improving working memory capacity can alter biochemical and neurophysiological function as well as, improve behavioral outcomes such as attention in a myriad of patient populations.

Below is a brief summary of Cogmed training studies that have been published in peer-reviewed journals in chronological order. Unless otherwise noted, the studies incorporate the working memory training program developed by Dr. Klingberg, which is the current basis of Cogmed training. Links to more extensive summaries of the different studies are included where available.

2002
Training of working memory in children with ADHD
Journal of Clinical and Experimental Neuropsychology
In the first published research on working memory training, Klingberg et al., 2002 conducted a two part, randomized, controlled study investigating the impact of training on: a) small sample of children with ADHD and b) healthy adults. Both children with ADHD and healthy adults improved compared to control groups on non-trained tasks of working memory, as well as on neuropsychological tests of inhibition and non-verbal reasoning. The study was the first indication that working memory capacity could be increased with intensive training. Limitations to this study include the small sample size, the lack of behavioral measures, and no long-term follow up.

2004
Increased prefrontal and parietal brain activity after training of working memory
Nature Neuroscience
This study by Olesen et al. was the first to examine brain changes following working memory training. Using fMRI, researchers documented training related changes in brain activity and improved working memory capacity in 3 healthy adults followed by a larger trial including 8 adults. Although the small samples highlight the need for replication, the findings are significant in that they provide the first demonstration of increased activity in the frontal and parietal lobes after working memory training.

2005
Computerized training of working memory in children with ADHD – a randomized, controlled trial
Journal of the American Academy of Child and Adolescent Psychiatry
In 2005, Klingberg et al. replicated the findings of Klingberg et al., 2002 in a randomized, controlled study of working memory training in 53 children diagnosed with ADHD. Children in the treatment group had significant gains in non-trained measures of working memory, non-verbal reasoning, and response inhibition compared to the control group. Although teacher ratings of reductions in ADHD symptoms failed to attain significance, significant reductions in parent ratings of ADHD symptoms were reported. The gains evident immediately after training were maintained 3 months post training.

2007
Computerized working memory training after stroke – a pilot study
Brain Injury
Westerberg et al. examined the impact of working memory training in 18 adult stroke victims who were randomly assigned to working memory training or to a no treatment (passive) control condition. Training resulted in significant improvement on non-trained measures of working memory and attention. Participants also reported decreased cognitive symptoms (ie., memory problems and attention lapses) and significant improvement in their daily functioning. The study is noteworthy because it suggests a potential role of working memory training in the rehabilitation of stroke victims.

Changes in cortical activity after training of working memory – a single-subject analysis
Psychology and Behavior
In 2007, Westerberg and Klingberg examined changes in brain activity after working memory training. Three healthy adults were scanned with fMRI while performing a working memory and a baseline task twice prior to and once after completing the standard 5-week Cogmed training program. Participants had improved working memory capacity following training as well as, significantly increased working memory-related brain activity in the middle and inferior frontal gyrus. Although this was a small sample study – not uncommon in fMRI research – this investigation provides additional evidence that working memory training impacts cortical activity and brain plasticity.

2008
Training and transfer effects of executive functions in preschool children
Developmental Science
This study by Thorell et al. examined the impact of visuo-spatial working memory training in 63 typically developing 4-5 year-old children. Participants were randomly assigned to either: a) adaptive visuo-spatial working memory training, b) computerized inhibitory control training, c) training with a commercially available video game with little or no impact on working memory or inhibition (active control) or d) no training (passive control). Pre-post intervention assessments revealed that only the working memory training group showed significant improvement in non-trained measures of auditory attention and both verbal and visuo-spatial working memory. Although the working memory training group did not improve on measures of inhibitory control or problem solving, this study showed that it was possible to enhance cognitive functions (ie., working memory and attention) in young children which are critical for academic success.

2009
Changes in cortical dopamine D1 receptor binding associated with cognitive training
Science
McNab et al. (2009) used positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) to examine the underlying biochemical impact of working memory training. Thirteen healthy volunteers participated in the standard 5-week Cogmed training and underwent assessment of working memory, fMRI and PET pre- and post-intervention. Participants not only increased working memory capacity but also, their improvement was correlated with decreased dopamine D1 receptor binding potential. Such changes in binding potential impact the availability of dopamine and are associated with changes in the density of cortical dopamine D1 receptors. Thus, this study builds upon prior research demonstrating brain activity changes following working memory training and is the first known demonstration that cognitive training alters fundamental aspects of brain biochemistry.

Adaptive training leads to sustained enhancement of poor working memory in children
Developmental Science
In 2009, Holmes et al. screened a large sample of children for working memory deficits. The lowest 15% were selected to participate with a sample of 42 children training with either adaptive or non-adaptive versions of Cogmed. Children receiving adaptive working memory training (ie., the standard Cogmed training protocol) showed gains on several non-trained measures of working memory that remained significant at 6 month follow up. Adaptive training participants also improved on an ecologically valid measure of instruction following and exhibited significant gains in a measure of math achievement 6 months after training. On the other hand, control participants (ie., non-adaptive training group) did not evidence significant gains on working memory measures post-training. This research was both the first independent study employing Cogmed training and the first to demonstrate that working memory training can produce gains on a standardized measure of academic achievement.

Working memory plasticity modulated by dopamine transporter genotype
Neuroscience Letters
Brehmer et al. investigated the influence of the DAT1 gene polymorphism (ie., a variation in a dopamine gene) on working memory plasticity, operationalized as performance gains across the course of working memory training. Twenty-nine young adults, ages 20 -31 years, were divided into two groups depending on whether they carried different versions of the dopamine gene (ie., the DAT 9/10 – repeat allele or the DAT 10- repeat allele) and each group of participants trained with the standard Cogmed training protocol. Post-training, DAT 9/10 carriers were found to exhibit greater working memory capacity gains than DAT 10 carries. It was posited that the gains evidenced by the DAT 9/10 carriers may have related to their having more active dopaminergic pathways. Importantly, this was the first study demonstrating a modulating effect of a genetic polymorphism on cognitive plasticity.

2010
Working memory deficits can be overcome: Impacts of training and medication on working memory in children with ADHD
Applied Cognitive Psychology
Holmes et al. examined the impact of medication treatment and Cogmed Working Memory Training on the working memory performance and IQ of 25 children with ADHD, 8-11 years old. Participants were assessed at 4 consecutive time points: T1 – prior to training, off medication, T2 – prior to training, on medication, T3 – after 5 weeks of Cogmed training, on medication and T4 – 6 months after training, on medication. The researchers found that medication treatment alone improved performance on only visuo-spatial working memory. However, when the children did working memory training in conjunction with medication treatment, they exhibited significant gains on visuo-spatial and verbal working memory tests and verbal and visual short-term memory tests. Although training had no significant impact on IQ scores, 6 month follow up data indicated that training gains in working memory had persisted. This research is noteworthy in that, Cogmed training led to working memory gains in children with ADHD beyond those provided by medication which persisted long after training had ended.

Working memory training for children with attention problems or hyperactivity: A school based pilot study
School Mental Health
In 2010, Children’s Hospital (Harvard University) researchers Mezzacappa and Buckner recruited 8 school children from a low SES background who had qualified for free breakfast and lunch at school. After training with Cogmed RM at school for five weeks, the children had improved significantly on measures of verbal and visuo-spatial working memory, as well as had decreased symptoms of ADHD as rated by teachers. This pilot study not only replicated findings that ADHD symptoms could be ameliorated by training but also, showed that Cogmed is equally as feasible an intervention for children from low SES environments.

Effects of working memory training on reading in children with special needs
Reading and Writing
Stockholm University’s Karin Dahlin investigated the efficacy of Cogmed training in improving reading comprehension in 57 Swedish children with special needs (ie., general learning problems and attention deficits). Following 5 weeks of Cogmed training, participants had improved on measures of visuo-spatial and verbal working memory, inhibition and most importantly, reading comprehension. Dahlin concluded that: “…training of working memory may be useful for children with reading comprehension problems, special-education needs, and attention problems” and posited that screening for working memory deficits would be valuable for identifying those at risk to struggle academically.

Training and plasticity of working memory
Trends in Cognitive Sciences
Although not a Cogmed training study, this review by Dr. Torkel Klingberg, cognitive neuroscientist from the Karolinska Institute, provides an overview of the psychological and neural correlates of working memory and working memory training. Dr. Klingberg summarizes the body of research investigating plasticity of working memory and outlines the use of Cogmed training for various populations with working memory deficits.

Computerized training of working memory in a group of patients suffering from acquired brain injury
Brain Injury
In 2010, Lundqvist et al. studied the use of Cogmed training in heterogeneous sample of 21 brain injured patients. All participants were struggling with working memory deficits resulting from a brain injury (ie., stroke, tumor, infection, TBI, hemorrhage) and were on average 3 years post injury. Participants were randomly assigned to either adaptive Cogmed training or a waitlist control group and assessed pre-, post- and 5 months following training. Following working memory training, patients evidenced improved attention, working memory, verbal inhibition, executive shifting and self-reported improved occupational performance. This study further solidifies Cogmed as a feasible intervention for stroke and brain injured patients and provides evidence of real world benefits following training.

A controlled trial of working memory training for children and adolescents with ADHD
Journal of Clinical Child & Adolescent Psychology
Beck et al. examined the efficacy of Cogmed training for 52 children and adolescents with ADHD and other co-morbid diagnoses. Participants, ages 7 – 17 years old, were recruited from a private school specializing in education for children with ADHD and other learning disabilities. Children were first screened for working memory deficits and then randomly assigned to either a training or waitlist control group. Post- training, there was a significant improvement in parent report of ADHD symptoms, executive functions, organization, and working memory. The researchers concluded that Cogmed training may be effective in improving core cognitive deficits underlying ADHD and thus, improve ADHD symptoms.

Computerized working memory training improves function in adolescents born at extremely low birth weight
The Journal of Pediatrics
Norwegian researchers Løhaugen et al. examined whether cognitive deficits experienced by children and adolescents who were born preterm with extremely low birth weight (ELBW) could be ameliorated with working memory training. Not only did ELBW children improve on non-trained working memory tasks following the Cogmed but also, the training effect generalized to verbal learning ability with gains remaining stable six months after training. The researchers concluded that Cogmed was an: “…effective intervention tool for improving memory and reducing core learning deficits in adolescents born at ELBW”.

Working memory training for children with cochlear implants: A pilot study
Journal of Speech, Language, and Hearing Research
Kronengerger et al. (2010) investigated the feasibility and efficacy of working memory training for improving memory and language skills in a sample of 9 deaf children with cochlear implants, ages 7- 15 years old. Although not significant at 6 month follow up, children did exhibit improved verbal and visuo-spatial working memory with improved parent report of working memory and attention up to one month post training. A noteworthy outcome in this study was that: “…almost all (89%) of the sample continued to show at least small improvement in Sentence Repetition scores at 6-Month Follow-Up, and nearly half (44%) of the sample showed large improvement in Sentence Repetition scores at 6-Month Follow-Up”. This improvement in sentence repetition is reflective of Cogmed’s positive impact on core underlying skills of language, memory and speech production which are deficit in the hearing impaired population.

2011
Component analysis of verbal versus spatial working memory training in adolescents with ADHD: A randomized, controlled trial
Child Neuropsychology
In a 2010 study, Gibson et al. investigated the utility of Unsworth and Engle’s dual-component model (2007) as a theoretical basis for the WM capacity differences in individuals with and without ADHD. Results from Gibson et al. (2010) suggested that maintenance of information in primary memory (PM) was largely intact in ADHD individuals and that secondary memory (SM) was the deficient WM structure in ADHD. Stemming from this 2010 study, Gibson et al. (2011) questioned whether Cogmed Working Memory Training was effectively training PM or SM or both in ADHD individuals. These researchers proposed that the Cogmed RM program, consisting of simple span tasks, was training only the PM component of working memory. Instead, complex tasks would be necessary to train the SM structure because these exercises force all but the last item of a span to be shifted from PM to SM and thus, the participant would be attempting to extract information from SM during training. Gibson et al. (2011) further reasoned that spatial simple span tasks may function like complex span tasks and may be more appropriate for training the SM component than verbal simple span tasks. Thus, one group of ADHD individuals trained with a spatial task-only version of Cogmed RM and another group of ADHD individuals with a verbal task-only version of Cogmed RM.

Gibson et al. (2011) found that when tested with immediate-free-recall tasks after training that there was no significant difference between the verbal and visuospatial Cogmed training groups. Participants in both groups recalled a greater number of items from PM post-intervention but, there was no improvement post-intervention in number of items recalled from SM. Based on these findings, Gibson et al., (2011) concluded that the spatial and verbal exercises in Cogmed are equally effective and that these simple span exercises primarily train the PM component of WM rather than the SM component. These researchers suggested that Cogmed should incorporate complex span tasks so to increase the likeliness that items to-be-remembered are lost from PM and thus, the user must try and extract these items from SM during training.

Gains in fluid intelligence after training non-verbal reasoning in 4-year-old children: A controlled randomized study
Developmental Science
In a double-blind, randomized, controlled study, Bergman-Nutley et al. investigated whether fluid intelligence could be improved by computerized non-verbal reasoning (NVR) training or working memory training. Preschool children were assigned to either: a) visuo-spatial working memory training, b) non-verbal reasoning training, c) combination working memory-NVR training or d) non-adaptive (placebo) combination training. Although only NVR training led to gains in fluid intelligence, working memory training still led to increased working memory capacity in preschoolers.

Preliminary evidence that allelic variation in the LMX1A gene influences training related working memory improvement
Neuropsychologia
The LMX1A gene impacts the proliferation, differentiation and maintenance of dopamine producing neurons in the brain. Bellander et al. investigated how individual differences (ie., allelic variations in single nucleotide polymorphisms(SNPs)) in the LMX1A gene influence one’s ability to benefit from working memory training. Twenty-nine younger adults (ages 20 -31 years) underwent genetic testing followed by Cogmed training. Although all of the participants improved in working memory measures post-training, individuals with the TT genotype for the rs4657412 SNP evidenced significantly greater gains in verbal working memory. This study underscores the importance of dopamine and dopamine-related genes in modulating the impacts of working memory training.

Neural correlates of training-related working-memory gains in old age
NeuroImage
Brehmer and colleagues investigated the relationship between behavioral performance and neural activity in 23 healthy, older adults (60-70 years) following five weeks of Cogmed training. Participants were randomly assigned to either adaptive Cogmed training or a non-adaptive placebo version of Cogmed. Using fMRI, all participants were measured before and after training doing a low difficulty working memory task and a high difficultly working memory task. Both groups evidenced decreased brain activity on the low and high difficultly working memory tasks. Compared to the placebo group, the adaptive training group had significantly larger decreases in memory and attention-related brains regions (frontal, occipital and temporal areas) as well as, subcortical increases when performing the high difficulty working memory task after training. The benefits of training thus unfolded in the context of more difficult working memory challenges as larger decreases in activation imply that the adaptive group needed to invest less neural energy. The adaptive group also improved over the control in measures of working memory, attention and episodic memory.

Working memory training for patients with acquired brain injury: effects in daily life
Scandinavian Journal of Occupational Therapy
Researchers Berit Johansson and Marjana Tornmalm at Habilitering & Hälsa, a rehabilitation facility in Stockholm Sweden, investigated the impact of Cogmed training on a heterogeneous sample of adults with acquired brain injury.
Participants also shared experiences of working memory deficits and learned about the function of working memory in everyday life for 30 minutes on each training day. Researchers collected self-report ratings of cognitive failures (CFQ; Cognitive Failures Questionnaire) and self-report ratings of occupational performance (COPM; Canadian Occupational Performance Measure) pre-, post- and 6 months after training. Participants kept records of their progress and problems with working memory training in a diary and reported subjective changes in daily functioning in an interview at 6 month follow-up. After training, participants reported reduced cognitive problems, improved performance at work and increased satisfaction at work. Although this study lacked a large sample size and control group, qualitative data of subjective changes in daily life were supported by data from the COPM and CFQ.

Conclusion
As demonstrated by the volume of published studies to date and by the breadth of topics covered, Cogmed has a strong body of literature supporting our claim to improve working memory and attention in people of all ages and with a variety of cognitive deficits. As a forerunner in the field of evidence based cognitive training, Cogmed has at its core a commitment to validation of our training solution through peer-reviewed, independent investigations. In addition to our published training studies, there have also been projects presented at national conferences and a wide range of ongoing studies, descriptions of which can be found in the research section of this website.