New Big Data Study Reveals Clear Link Between "Dose" of Brain Training and Improvements in Cognition Across All Ages

A study led by psychologist Allen Osman, Ph.D., and the research team at Lumos Labs, leverages big data from Lumosity gameplays to shed new light on the impact of cognitive training on cognitive performance. The study, published in PLOS ONE, aimed to address the limitations of several prior studies by performing a large-scale analysis on a real-world cohort of over 100,000 individuals. The results indicate a strong, consistent relationship between the volume of training and significant performance improvements across multiple cognitive domains for all age demographics. In addition, the findings suggest that a lack of improvement from cognitive training seen by some studies might be attributable to insufficient training dosage.

Dr. Osman elaborates, "The excitement around brain training stems from its possible use to enhance normal cognition or help ameliorate disorders of cognition, but prior research has led to mixed findings. What this study does really well is to not only show that a dose-response relation is ubiquitous across the adult lifespan and multiple cognitive domains, but also that some earlier failures to find effects from brain training may have been due to small, insufficient doses."

The study leveraged a massive anonymized data set from Lumosity users across all demographics who also undertook a separate set of neurocognitive tests at multiple timepoints throughout their training. The Lumosity program, consisting of a library of cognitive training tasks adapted into brief games, categorizes these games into several cognitive domains, including Memory, Attention, Problem Solving, Cognitive Flexibility, Information Processing Speed, Language, and Math.

Individuals in the study also completed the NeuroCognitive Performance Test (NCPT), which includes validated, computerized neurocognitive tests assessing short term memory, working memory, information processing speed, reasoning, and other cognitive domains. The NCPT tests, not gamified, are separate from the tasks used in Lumosity games. The study monitored NCPT performance changes over the course of Lumosity training, allowing researchers to measure "transfer of learning"—the extent to which training on Lumosity games transfers to performance on distinct cognitive tests.

Dr. Osman notes that transfer of learning from Lumosity was already demonstrated by prior studies, including a large, randomized, active-controlled trial from Hardy and colleagues (2015). "But unlike in a traditional trial, our 'big data' approach allowed us to get a precise measure of the 'dose-response function' between training and test performance; this is the way that different amounts of training correspond with different sized improvements in neurocognitive performance."

These dose-response functions were consistent across different cognitive domains and diverse groups based on age, gender, and educational level. More training invariably led to greater NCPT improvements, and a substantial fraction of the maximal improvement could be attained with about 20 hours of cumulative Lumosity training.

The study also sheds further light on whether the effectiveness of cognitive training changes across the adult lifespan. Interestingly, the same amount of learning transfer was observed across all age groups, implying the efficacy of cognitive training did not decrease with age—a positive indication for older individuals engaging in cognitive training. This finding replicates similar ones by Ng and colleagues (2021), who compared the effectiveness of cognitive training for older vs. younger adults in the controlled experimental study performed by Hardy and colleagues (2015).

Finally, the study offers some intriguing clues about mechanisms of learning and cognitive aging. Performance improvements on the NCPT could stem from either the transfer of learning from cognitive training or simply from taking the NCPT multiple times. The analytical approach taken by the authors allowed them to differentiate between improvements attributed to cognitive training versus direct practice of the NCPT. Analysis of the data revealed that while older adults saw less gain from direct NCPT practice when compared to younger individuals, transfer-related improvement was consistent across all ages. Dr. Osman explains, "The dissociation between direct practice and transfer suggests that these two forms of learning did not involve identical cognitive processes and that those involved in cognitive training effects were preserved across the adult lifespan."

The article, entitled "Transfer of learning: Analysis of dose-response functions from a large-scale, online, cognitive training dataset," is Open Access and available online on the PLOS ONE website.

About Lumos Labs

Lumos Labs, the creator of Lumosity, is a pioneer in cognitive health and digital medicine with the mission of empowering everyone to discover what their mind can do. The company's Lumosity program has attracted over 100 million members from 180+ countries, contributing to a dataset of over 7.5 billion cognitive games played, which serves to support over 100 collaborations with university research groups. Lumosity DTx, the Lumos Labs medical division, is at the forefront of developing FDA-cleared digital medicine. Awarded by the Future of Privacy Forum for Data Stewardship, Lumos Labs prides itself on supporting the advancement of cognitive research while protecting the privacy of its members.

More information:
Presented Study:
Osman, A. M., Jaffe, P. I., Ng, N. F., Kerlan, K. R., & Schafer, R. J. (2023). Transfer of learning: Analysis of dose-response functions from a large-scale, online, cognitive training dataset. PLOS ONE, 18(5), e0281095.

Additional References:
Ng, N. F., Osman, A. M., Kerlan, K. R., Doraiswamy, P. M., & Schafer, R. J. (2021). Computerized Cognitive Training by Healthy Older and Younger Adults: Age Comparisons of Overall Efficacy and Selective Effects on Cognition. Frontiers in Neurology, 11, 564317. doi.org/10.3389/fneur.2020.564317

Hardy JL, Nelson RA, Thomason ME, Sternberg DA, Katovich K, Farzin F, et al. (2015) Enhancing Cognitive Abilities with Comprehensive Training: A Large, Online, Randomized, Active-Controlled Trial. PLOS ONE 10(9): e0134467. doi.org/10.1371/journal.pone.0134467
Press Contact: Seth S Gourson—seth.gourson@lumoslabs.com