Questions & Answers: Training of working memory in children with attention deficits

Presented by:
Torkel Klingberg, M.D., Ph.D., Karolinska Institute

Prepared by:
Kathryn Ralph
Research Coordinator
Cogmed
Pearson Clinical Assessment

Below please find answers to questions asked during the live webinar:

1. How does someone initiate the Cogmed training? Is it covered by insurance?

Cogmed can be found across the world. To find a Cogmed provider near you click here. If you are a healthcare professional and you are interested becoming a Cogmed practitioner click here. If you are an educator and you are interested in offering Cogmed in your school click here.

2. What was the typical frequency and duration of WM training sessions?

Cogmed working memory training for school-age children and adults is 30 to 40 minutes per day, 5 days per week for 5 weeks. Cogmed for preschoolers is 15 minutes per day, 5 days per week for 5 weeks.

3. What is the age range for participation in Cogmed training program?

People of all ages have used Cogmed clinically and in research. There are three versions of the Cogmed program: Cogmed JM for pre-schoolers, Cogmed RM for children and Cogmed QM for adults.

4. Has training of phonological loop been studied to see if you can improve non-spatial verbal working memory?

Cogmed has both visuo-spatial exercises and verbal (phonological) exercises. There are more visuo-spatial exercises because evidence suggests that visuo-spatial training leads to improvements in both visuospatial and verbal working memory. In 2009, Thorell et al. reported: “An interesting finding of the present study was that, unlike Klingberg et al. (2002, 2005), our training program only included tasks of visuo-spatial WM. Thus, there was a transfer of effect of visuo-spatial training to the verbal domain of WM, which is in line with previous neuroimaging findings showing evidence of supramodal WM areas (i.e. areas that are active irrespective of the type of stimuli being held in WM) within the parietal and prefrontal cortex (Curtis & D’Esposito, 2003; Hautzel, Mottaghy, Schmidt, Zemb, Shah, Muller-Gartner & Krause, 2002; Klingberg, 1998).”

5. Should Cogmed trainees be explicitly using strategies during training.

Cogmed is not designed to teach explicit strategies and users should therefore not employ explicit strategies during training. Instead, Cogmed as an adaptive, working memory training program is intended to increase working memory capacity by requiring the user both to hold in mind and manipulate an increasing number of items over time. Unlike strategy use, which is material-specific, working memory training gains generalize to enhancements in other domains. Holmes, Gathercole and Dunning (2009) explain, “There has been some success in boosting performance on WM tests through strategy training (Turley-Ames & Whitfield, 2003), although because gains often do not extend beyond trained tasks they are unlikely to yield substantial benefits for the many and varied learning situations in which children depend on WM”.

6. Do we have any evidence yet about the merits of the Extension Training (100 additional sessions over the course of a calendar year)?

In response to customer requests for continued access to Cogmed after completing the 25 days of working memory training, Cogmed created Extension Training. For up to one year, Extension Training allows the user to complete 100 additional Cogmed sessions with 4 exercises per session. There is no coach support for these additional sessions and there is also no research basis for this additional training. Future investigations will look at what training time and protocol yield optimum benefits and whether booster sessions are effective. We do know however, that after 25 days of sustained working memory training, users show improved working memory, attention and behavioural outcomes.

7. Do children enjoy the training? Is it difficult to convince them to persist with it? Do parents or clinicians report any improvement in behavioural functioning (i.e., frustration tolerance, homework completion, improved social skills) after training?

In North America, 90% of children who start Cogmed are compliant, completing at least 20 training days in 8 weeks. In addition to the engaging interface of Cogmed RM and the reward game RoboRacing that is available at the end of each training day, the coaching protocol ensures that children stay motivated over the 25 sessions. Not only do children enjoy training but also, parents frequently report changes in their child’s everyday functioning and behaviour. Read the Cogmed user stories.

8. I believe I read Cogmed helps 80% of clients. How much improvement can parents expect to see in their child after five weeks? During or after training, is this documented in everyday functional life skills?

Improvements after Cogmed are dependent on a variety of factors including the initial deficit of the user, the quality of their training (i.e. compliant vs. non-compliant) and appropriate monitoring of change (i.e. assessment). However, data from Cogmed clinical practices and academic research (Klingberg et al., 2005) consistently shows that 80% of users have improved behavioural symptoms after training. Generally, training gains emerge in the weeks and months following training and a 6 month follow up a mandatory coaching requirement in the clinical setting for this very reason. In research gains were also substantiated after training in Klingberg et al. (2005) where 60% of participants in the treatment group who fulfilled criteria for ADHD symptoms prior to intervention no longer met these criteria post-intervention.

9. If my client fell in the 20% that showed no significant change, how do I justify the cost of 25 sessions without the desired result?

Before an individual begins Cogmed training, it is vital for them to know that 20% of users do not report experiencing significant benefits. It is the responsibility of the Cogmed Coach to discuss with the user the nature of their deficit, the need for realistic expectations during and after training and the importance of continued assessment after training. There is a strong body of research supporting the use of Cogmed to alleviate working memory constraint and it is recommended that Cogmed practitioners understand and explain the research basis for working memory training.

10. Have you been able to quantify or figure out the characteristics (i.e. co-morbid disorders) of which children may not benefit from Cogmed (i.e. who are those 20%)?

Although a large number of Cogmed users improve after training (80%), it is still unclear why some Cogmed users do not experience benefits. Cogmed actively supports over 50 research projects by academic researchers whose data may shed light on this question. The Cogmed team has also begun analyzing de-identified clinical data from practices all over the world in order to further clarify for what populations Cogmed works best.

11. Are there any school-based studies planned?

There are number of ongoing research projects taking place in schools. One such study is investigating the impact of Cogmed compared to the Bookworm reading intervention program. Read more. Dr. Joni Holmes has also recently completed data collection with three different groups of children who trained either before or after school. Read more. For a poster of her preliminary findings click here. Further, numerous published Cogmed training studies have been implemented in the school environment including: Holmes et al., 2009, Thorell et al., 2009,and Mezzacappa and Buckner, 2010.

12. Have you compared medication to training efficacy effects without medication as well as follow-up results?

In 2010, Holmes et al. evaluated the working memory of a sample of children with ADHD while off medication, on medication alone and on-medication during and after Cogmed training. Read more. To date, there have been no controlled studies looking at the impact of Cogmed vs. medication.

13. Can you say with any degree of certainty what the likelihood of reducing/eliminating medication (i.e. psychostimulants) would be with WM training?

No, there is not enough clinical or research evidence to make statements about reducing or eliminating medication after Cogmed training.

14. Has there been any data that would raise concerns as to whether there is any suppression of the training effect with the concurrent use of medication?

No, there is no published data raising concerns that medication might suppress the training effect. To the contrary, Holmes et al. (2010) found that ADHD children training with Cogmed while on psychostimulant medication had significantly improved working memory compared to their working memory on medication alone. It should be noted that the standard protocol calls for users to be stabilized on their medication before training.

15. Have there been studies about whether the training would show improvement in attention problems of patients with emotional symptoms?

Yes. Dr. Julie Hadwin at the University of Southampton has recently completed a project investigating the impact of Cogmed on children with social, emotional and behavioral difficulties. Read more. Dr. Hadwin is following up with a second project looking at anxiety and school performance. Read more.

16. Is WM training useful for cognitive bias modification for anxiety and addiction?

There has not been any research conducted using Cogmed as a part of cognitive bias modification.

17. The WISC has a working memory subtest. Does the training increase the WISC scores?

The WISC III and WISC IV are frequently used assessments in Cogmed training studies. In a study of ADHD children by Klingberg et al. (2005) and another study of special needs children by Dahlin (2011), the digit span from the WISC-III was used to assess verbal working memory changes and effect sizes were 0.59 and 0.66 respectively. Lohaugen et al. (2010) used the Working Memory Index from the WISC-IV in their study of adolescents born at extremely low birth weight and found significant improvements on digit span, letter number sequencing and spatial span at post-test and 6 month follow up. Mezzacappa and Buckner assessed ADHD children using the digit span from the WISC-IV and observed an effect size of 0.93.

18. I’ve been asked about the similarity between some of the Cogmed exercises and the digit span subtest of the WISC cognitive test and its possible impact regarding teaching to the test. How would you respond to this?

One method that is regularly employed by researchers conducting Cogmed training studies is to assess users with both trained and non-trained working memory tasks. A trained task is a task that has been practiced for the duration of training and thus, one would expect that the user would improve on this task. To demonstrate gains beyond task-specific improvements, non-trained tasks assess if there has been near-transfer by testing for improvement in the same construct that was trained (i.e. working memory) but with tasks differ in presentation, design and response mode (Bergman Nutley et al., 2011).

For example, a Cogmed user trains with a computerized exercise where they must click a mouse to repeat the order in which lamps light up on a 4×4 grid (trained task). Then, the user is assessed with a span board task where a researcher points to wooden blocks in a random order and the user must repeat the sequence by pointing out the block with their hand (non-trained task). After Cogmed users improve not only on trained but also, on non-trained working memory tasks (Klingberg et al., 2002, 2005, 2010; Westerberg et al., 2007; Thorell et al., 2009; Dahlin, 2011)

Other cognitive tasks and rating scales are also used to assess far-transfer that is; generalization of training effects to constructs beyond working memory such as attention and behaviour. For example, Westerberg et al. (2007) found that adult stroke survivors improved on the span board and digit span tests (working memory; near transfer), on the PASAT (attention; far transfer), as well as on the Cognitive Failures Questionnaire (self report of cognitive failures; far transfer).

19. Would you expect that single word reading would not improve with working memory training, but comprehension of longer or more complex sentences and paragraphs should?

In a controlled study of school-aged children who completed Cogmed training by Holmes et al. (2009), no significant gains were found in basic word reading, as measured by the Wechsler Objective Reading Dimensions (WOND), a sub-test of the Wechsler Individual Achievement Test (WIAT). The absence of benefits on the reading measure may reflect the fact that the test used – a basic word reading test – does not depend heavily on working memory skills. However, there is preliminary evidence that Cogmed training can enhance reading comprehension. In a study of special needs students (Dahlin, 2011), children who did Cogmed training improved in their ability to read narrative texts and to answer questions based on those texts both immediately after training and at 6 month follow up.

20. What is the impact of Cogmed WM training on executive functions?

In addition to improvements in visuo-spatial and verbal working memory, researchers conducting Cogmed training studies with  adult stroke survivors (Westerberg et al., 2007), adults with acquired brain injury  (Lundqvist el al., 2010) and typically aging older adults (Brehmer et al., 2011) have observed a statistically significant  transfer of the training effect to attention, as measured the Paced Auditory Serial Addition Task (PASAT).

Participants in Lundqvist et al., 2010 also evidenced improved verbal inhibition and executive shifting, as measured by the Color Word Interference Test – condition 4-Inhibition/Switching (CWIT). Improvements on the Stroop task (inhibition of a pre-potent response) were also observed in children with special needs (ADHD/LD) (Dahlin, 2011),  two training studies of children with ADHD (Klingberg et al., 2002; 2005) and in young adults (Olesen et al., 2004). However, such improvements in inhibition were not seen in adult stroke survivors (Westerberg et al., 2007) or typically functioning older adults (Brehmer el al., 2011). Thus, it is unclear whether improvements in inhibitory function after training are specific or related generally to top-down attention. To read about a study comparing working memory training to inhibition training in pre-school children, click here.

The impact of working memory training on non-verbal reasoning has also been varied.  Improvements were not seen in typically aging older adults on Raven’s Progressive Matrices (Brehmer et al., 2011). In a 2011 study, Bergman-Nutley et al. compared pre-school children in a  working memory training group to children in a non-verbal reasoning training group and found improved working memory but, no significant gains in non-verbal reasoning for the working memory training participants. However, ADHD children improved on Raven’s Colored Progressive Matrices (Klingberg et al. 2002; 2005) as did typically functioning younger adults on Raven’s Advanced Progressive Matrices (Olesen et al., 2004; Klingberg & Westerberg, 2007)after Cogmed training. It has been posited that these inconsistent findings may be due to difference in populations and the tasks used to assess transfer effects.