Study: 12 Weeks of CogniFit Training Improved Global Cognition in Adults Over 60

What the study tested

Roughly 55 million people live with dementia worldwide, with nearly 10 million new cases each year, per WHO figures cited in the paper. Mild cognitive impairment, a preclinical stage between normal aging and dementia, affects an estimated 15% to 23% of community-dwelling adults aged 50 and over, and predicts progression to dementia in up to 46% of cases within 3 years. With the global population over 60 projected to more than double from 1 billion in 2020 to 2.1 billion by 2050, the researchers set out to test whether a digital, nonpharmacological intervention could help maintain cognitive performance in healthy older adults.^[1]

The team at the Universidad de Castilla-La Mancha ran a quasi-experimental pretest–posttest study with a control group from October 2022 to May 2023, following the TREND Statement recommendations for nonrandomized studies. Of 74 people who expressed interest, 60 began the study and 45 completed it (75%): 23 in the intervention group and 22 in the control group, with similar completion rates (76.6% vs. 73.33%). Participants were community-dwelling adults over 60, mean age 70.45 years in the intervention group and 71.5 in controls, 64.4% women, recruited through neighborhood associations and senior organizations in Spain. All were fully autonomous in daily activities, with no diagnosed neurocognitive disorders.^[1]

The intervention group trained at home on the CogniFit app: 36 sessions over 12 weeks, three sessions per week, each lasting 20 to 30 minutes. The program began with CogniFit's computerized neurocognitive assessment (about 40 minutes, internal consistency α = .930 in this sample) and personalized the difficulty of each exercise after every session. It targeted nine abilities commonly affected in old age, including working memory, naming, visual and spatial perception, processing speed, hand–eye coordination, and response time. The control group received no intervention.

Outcomes were measured at baseline and again 8 weeks after the intervention ended, a lag designed to separate lasting improvement from short-term practice effects. The battery combined the CogniFit assessment, the 30-point Montreal Cognitive Assessment (MoCA), the 28-item Memory Failures of Everyday (MFE) questionnaire, the EuroQol EQ-5D-3L quality-of-life scale, the Oviedo Sleep Questionnaire, and the GDS-15 depression scale.

What it found

The cognitive gains were broad, not narrow. Compared with controls, the trained group improved significantly in 13 cognitive domains, including naming (p < .001), working memory (p = .007), visual perception (p < .001), hand–eye coordination (p = .005), response time (p < .001), contextual memory (p < .001), short-term memory (p < .001), and monitoring (p < .001). Within the intervention group, 16 domains improved significantly from baseline.^[1]

The result was corroborated by an instrument independent of the training platform: on the MoCA, trained participants rose from a mean of 25.55 (± 3.00) to 26.95 (± 2.01), crossing the test's threshold of 26 for normal cognition, while the control group slipped from 26.1 to 25.05. The between-group difference was significant (p = .044).^[1]

On everyday memory, the pattern was protective rather than enhancing. MFE memory-failure scores in the trained group went from 11.9 to 11.4 points (a nonsignificant change, p = .336), while controls deteriorated from 15.59 to 18.68 points (p = .003). Training did not erase memory lapses, but trained participants avoided the measurable decline that controls experienced over the same months.^[1]

Adherence was strong: every participant in the intervention group completed at least the prescribed 36 trainings, and the mean was 46.87 sessions (± 13.775), about 30% more than required. In post-intervention correlations, a higher CogniFit general score was associated with higher MoCA scores (rho = .332), fewer everyday memory failures (rho = −.404), and lower insomnia scores (rho = −.404).^[1]

Secondary health outcomes were mostly flat. There were no significant improvements in depression, sleep, or quality of life in the trained group; both groups deteriorated somewhat in quality of life over the study period. The authors are direct:

Limitations, what this study does not show

The authors list several constraints:

No randomization. Participants were allocated to groups by availability and interest, not chance. The quasi-experimental design carries threats to internal validity such as nonrandom selection and maturation effects, mitigated, but not eliminated, by validated instruments and a nonequivalent control group.

Small sample and attrition. Only 45 of 74 initially interested people completed the study, and the baseline CogniFit score was higher in the intervention group than in controls. Voluntary participation may also bias the sample relative to the general population, and the small sample makes internal-consistency coefficients (Cronbach's alpha) imprecise for some instruments.^[1]

Self-reported secondary measures. Sleep quality and quality of life relied on participant self-report. The headline finding is bounded: this study showed no significant gains in quality of life, sleep, or depression, the benefits were cognitive. The follow-up window was 8 weeks after training ended; longer-term durability was not tested.

Where this fits in the broader evidence

This is the first study of its kind in Spain and, per the authors, the first to test whether CogniFit-driven cognitive improvement extends to memory failure in everyday life. It builds on a consistent prior record: a 2022 systematic review by Nguyen et al. of commercially available brain-training programs found that, of the seven studies included, the only tools that yielded significant results were BrainHQ and CogniFit. Earlier trials reported CogniFit-related gains in working memory and processing speed in older adults (Gigle et al., 2013; Peretz et al., 2011), improved sleep quality and cognition in older adults with insomnia (Haimov & Shatil, 2013), and cognitive benefits in older diabetic adults at higher dementia risk (Bahar-Fuchs et al., 2020). The 2026 Spanish study extends that line with a between-group cognitive result measured 8 weeks after training ended, plus the novel everyday-memory protection finding. See the full CogniFit research index for all published studies.