Do brain games and apps really work and improve memory?

Answer:

Brain training apps (computer or mobile phone applications and games) promise to improve your memory and attention skills, ability to make good decisions, and even vision and hearing. There are numerous products on the market (Brain HQ, Lumosity, Cogmed, Project: EVO, etc.) and the claims of these companies are tantalizing: Who wouldn't want to get smarter by playing a game on their phone? While many of these products originate from scientific research that has shown positive effect in some individuals, the effectiveness of these apps are highly controversial: Scientists have noted the limitations of many of these products and the Federal Trade Commission has taken action against multiple brain training companies on the ground that their advertising practices are often misleading. At issue is that the research behind these apps is early stage and there is little understanding of which apps may provide benefits for a given individual. Thus while use of some of these apps may benefit some individuals, there is little differentiation in the marketplace and it is difficult to make informed buying decision.

What they are and how they're used
Brain training programs, or "brain games," typically involve practice with gamified versions of tasks used in neuropsychological examinations or psychology research. Standard tasks include memory span tasks (e.g. how many items can your remember), vision tasks (e.g. identifying stimuli that are briefly flashed in your peripheral vision or seeing how many moving objects can you track simultaneously), executive function tasks (e.g. can you sort objects when the sorting rules keep changing, can you report the direction of an arrow when it is surrounded by other arrows that are pointing a different direction, or how quickly can you stop a response to a target that you have already initiated?), among many others. These tasks have been adapted for use on computers, phones and tablets, and in some cases game consoles. 

Most companies provide a large selection of tasks to choose from and suggest training regimes where you mix and match tasks. However, most of the basic research that these products are based upon has come from studies where people practice a single task for an extended period of time (often 10 hours or more).

The evidence
Most research of brain training originates from early stage university-based research studies, where numerous groups have found promising results that certain procedures can give rise to improvements in tests of cognition in research settings. Procedures used in commercially available brain training apps are typically variants of these university-based proof of concept studies. For example one of the most common procedures is training with the "n-back" task, where participants are required to indicate whenever a current stimulus matches the stimulus that was presented earlier (i.e., n-items back) in the sequence. Although some studies have found training benefits to transfer to other memory tasks and even some intelligence tasks, training benefits are highly varied across studies and the extent to which n-back training transfers to real-world is still heavily debated. In fact, a study published in 2018 found that even though brain training led to modest improvements on an n-back tasks, these improvements did not transfer to related tasks, on which the participants scored no better than people who had not had brain training. The researchers concluded that "brain training protocols that focus on increasing the capacity of short-term memory do not yield generalizable improvements to cognition, regardless of the specific training task employed" (Stojanoski, Neuropsychologia, 2018).

The largest and best controlled study is the Advanced Cognitive Training for Independent and Vital Elderly (ACTIVE) trial, which started in 1996 with about 2,800 participants aged 65 and over who were randomized to three active training conditions: memory training (strategies to improve episodic memory), speed of processing training (focused on searching for briefly presented visual patterns), and a reasoning training (solving problems containing a serial pattern). Each training had a matched control group that did not receive the training. At five years and 10 years out, subjects in all three intervention groups reported significantly less difficulty than did participants in the control group in performing instrumental activities of daily living (Rebok, J Am Geriatr Soc 2014). In addition, older drivers who completed cognitive speed of processing training were 40% less likely to cease driving over the subsequent three years (Edwards, J Gerontol A Biol Sci Med Sci 2009).

Products
Currently the speed of processing training is being sold by Posit Science as part of their BrainHQ product. Notably, the commercial version of speed of processing training differs from that used in the ACTIVE trial and with little understanding of the active components that lead to effective brain training, it is unclear whether the commercial offering is more or less effective than the clinically-tested original.

Products with more preliminary clinical evidence include the following -- but it will take time before these results and products are fully vetted by the scientific and medical communities:

• Lumosity (Lumos Labs) has apparently shown greater improvements on a number of cognitive outcome measures than has playing crossword puzzles. 


• Cogmed (Pearson) may help treat attention and memory deficits in children with ADHD and seniors with cognitive decline, and it may enhance school and workplace performance in normally functioning individuals. 


• Project: EVO (Akili Interactive) is claimed to have shown clinical efficacy in children with ADHD. 

Cost
Brain training apps vary greatly in price. Many apps are freely available, commercial apps (such as Lumosity and BrainHQ) cost less than to $100 a year, and others that require the involvement of clinicians, such as Cogmed can cost thousands of dollars. There is little evidence that the more expensive products provide better results than those that are less expensive.

The bottom line:
The promises of brain training apps are often exaggerated; however, this is an active research area where there is significant promise that brain training can provide benefits. Further, there are no known harms of brain training programs, other than loss of time and money. Consumers need to understand that research is still in progress and that the active components of these products are not well understood and products that look very similar may have very different outcomes. Thus, we recommend a wait and see approach, consumers should carefully evaluate products in this space and try them out with guarded expectations.