Review of: Hill N et al, Am J Psychiatry 2017;174(4):329–340.

Study Type: Systematic review and meta-analysis of randomized controlled trials

Can patients train their way to better cognition? We’ve all seen the ads from companies such as Lumosity implying that fun, computer-based learning games will help your mind work better. Such methods are called computerized cognitive training (CCT), and past systematic reviews have had weaknesses, such as combining CCT with other interventions or including studies that were not randomized controlled trials. These researchers performed a systematic review that was more rigorous.

Methods
The study authors located 29 studies that met their strict criteria for inclusion in the meta-analysis. All studies had to randomly assign people to both a treatment arm and a control arm, and the CCT interventions had to include at least 4 hours of video games or virtual reality games. In almost all of the trials, the training was given in a group format with supervisors overseeing the process. The studies tested many commercial and nonprofit products, including Lumosity, BrainFitness, CogniPlus, Sociable, and others.

In all, 17 studies enrolled patients with mild cognitive impairment, or MCI (N = 686, CCT:N = 351, mean age = 67–81), while 12 studies enrolled patients with dementia (N = 389, CCT:N = 201, mean age = 66–81). Outcome variables included changes in cognition scores, activities of daily living, and psychosocial functioning. The length of these studies was relatively short, often lasting 1 to 3 months, with some going a little longer.

Results
CCT bested control groups for patients with MCI on several measures, with an overall effect size on cognition of 0.35. CCT training led to moderate improvements in most domains, including verbal memory, nonverbal learning, working memory, and attention. Interestingly, there were also improvements in depression and quality-of-life measures. However, CCT-exposed patients did not improve in measures of executive ­functioning.

CCT was less helpful for patients with dementia. The few significant results hinged on two studies looking at non-traditional, highly stimulating varieties of CCT—virtual reality and Nintendo Wii.

TCPR’s Take
The results appear promising for patients with MCI, and to some degree even for those with dementia. However, the follow-up was short (1 to 3 months), and it’s not yet clear whether such cognitive gains would be maintained once patients stop the training. In addition, these programs were supervised group programs, whereas most of the heavily advertised commercial products that your patients are likely to choose are home-based and oriented toward individual use.

Practice Implications
Tell your patients with MCI that computerized training programs might be helpful for memory improvement, but that the best evidence is for organized group programs with trained supervisors. Home-based single-player computer games might be helpful, but we need more evidence before giving them a strong endorsement.