Review: 7 Brain-Training Programs Were Graded for Healthy Aging, What the Evidence Showed, and What It Did Not

What the review tested

The paper cites a worldwide computer-based brain-fitness market estimated at $295 million in 2009, a rise of nearly 35% over 2008.^[1] Against that backdrop, the authors set out to test whether the validation literature behind these consumer products actually supported claims that they help maintain brain health in cognitively healthy older adults aged 50 years or older.

A team based at the McCusker Alzheimer's Research Foundation and three Western Australian universities searched PubMed and Google Scholar and each program's website, with the search updated through September 2015.^[1] Eligible studies were English-language, peer-reviewed clinical trials of computerized training in healthy older adults with a formal cognitive outcome measure; conference abstracts, video-game training, and studies of clinical populations were excluded.

Two reviewers independently screened titles and abstracts and extracted study source, sample size, age, intervention duration, intensity, and outcomes. Methodological quality was rated with the Physiotherapy Evidence Database (PEDro) scale (maximum 10; >6 = high, 5–6 = moderate, <5 = poor). Programs were then graded against an evidence-level scheme adapted from Cicerone et al.: Level I required at least two well-designed randomized or quasi-randomized trials (one high, one at least moderate quality); Level II required one high-quality RCT; Level III rested on moderate-to-poor trials or other designs.

What it found: an evidence-quality ladder

The screening flow narrowed an initial 32 programs to 18 with potentially eligible trials, then to 7 programs with usable clinical studies; 11 programs were removed for having no empirical evidence at all. After duplicates were removed, 7,985 records remained; 244 full-text articles were assessed, and 26 supporting studies (including follow-ups) were reviewed.^[1]

Quality was a relative strength of this evidence base. The mean PEDro score across the 18 trials without follow-ups was 7.05, with 12 studies rated high, 6 moderate, and none poor. The review sorted the 7 programs into three tiers:

For the Level I programs, the review reported that across the 13 studies in that tier, 7 studies showed improved processing speed, 3 reported improved attention (with small, medium, or large effect sizes), and 5 studies reported improved memory or working memory with small-to-medium effect size; only 2 studies showed significantly improved executive functions with small effect size. The single largest and most-cited trial in the review, the ACTIVE trial, recruited 2,832 participants aged 65 or older and found benefits in reasoning and speed of processing that, with occasional booster sessions, appeared to be sustained over 2, 5, and possibly 10 years.^[4]

Where CogniFit sits in this review

Cognifit was one of only two programs the review placed at Level I, alongside Posit Science. Cognifit interventions were evaluated in 3 studies, 1 high-quality and 2 moderate-quality on the PEDro scale.^[1] One was a randomized, double-blind trial in 155 healthy older adults reporting that, while both the training and an active control improved cognition, the training was more effective for visuospatial working memory, learning, and attention.^[3] Another randomized study in 118 healthy older adults used a four-condition design comparing physical activity, brain training, a combination, and a control group; the brain-training and combined groups showed enhanced memory and processing speed including hand-eye coordination.^[2]

Across that trial, the review noted, the combined-group benefit was attributed by the original authors to the brain-training component rather than the physical-activity component. Per the review, the Cognifit studies covered memory, processing speed, attention, executive functions, and visuospatial working memory.

Limitations, what this review does not show

The authors are candid about the boundaries of what they could conclude:

Funding and selective-reporting bias. Of the studies reviewed, 12 were conducted or funded by the program's own company and 6 were independent. Company-funded studies carried a higher average PEDro score (7.4) than independently funded ones (6.3). The abstract states plainly that "caution must be taken regarding any potential bias due to selective reporting."

It is not a meta-analysis. The review classified programs by evidence level and quality but did not pool effect sizes into a single estimate, and it does not establish that training prevents dementia. The authors note that cognitive training "may provide benefits in preventing/slowing progression of cognitive decline," but "it may not necessarily reverse the disease trajectory."

Programs are hard to compare head-to-head. Training doses, durations (ranging from 4 weeks to 16 weeks for Level I studies), control conditions, and outcome batteries differed widely, so the review concluded it is "difficult to compare any two programs side by side."

Where this fits in the broader evidence

The review's overall verdict is measured rather than promotional. The authors conclude that "current evidence supports that at least some commercially available computerized brain training products can assist in promoting healthy brain aging," while stressing the funding caveat and the gap between improving test scores and altering disease course.^[1] It complements later evidence maps of the same field, such as a 2024 scoping review of smartphone and web-based brain-training apps in older adults, and trial-level evidence such as a 2026 Spanish study of cognitive training in adults over 60 and a comparison of cognitive training versus aerobic exercise. Browse the full list of published CogniFit research on our research studies index, or see condition-specific evidence pages such as cognitive training in multiple sclerosis and cognitive training studied after stroke.