Brain & Cognition Last reviewed: June 30, 2026

Does cognitive training prevent or delay dementia?

Weak Evidence
Confidence Score 35%
โŒ

NO

Limited evidence suggests computerized cognitive training may produce modest improvements in trained cognitive domains, but claims of dementia prevention are not well-supported. Transfer to real-world function and long-term neuroprotection remain unproven.

The Verdict

Limited evidence suggests computerized cognitive training may produce modest improvements in trained cognitive domains, but claims of dementia prevention are not well-supported. Transfer to real-world function and long-term neuroprotection remain unproven.

What the Evidence Shows

The question of whether cognitive training can prevent or delay dementia has generated intense debate. Lampit et al. (2014) conducted a meta-analysis of 52 randomized trials including over 4,800 healthy older adults and found that computerized cognitive training produced small but significant improvements in cognitive function (overall effect size g = 0.22). However, benefits were largely restricted to the trained domain with limited transfer to untrained tasks or daily functioning. The ACTIVE trial, the largest and longest-running cognitive training RCT, showed that after 10 years of follow-up (Rebok et al., 2014), trained groups maintained cognitive advantages in their specific trained domain (memory, reasoning, or processing speed) compared to controls. Edwards et al. (2017) reported that specifically the speed-of-processing training arm of the ACTIVE trial was associated with a 29% lower risk of incident dementia after 10 years. However, this was a secondary analysis of a single training type and has not been independently replicated. Critically, Simons et al. (2016) published a comprehensive consensus statement from the scientific community arguing that claims made by brain training companies far exceed the evidence. They found that while training can improve performance on trained tasks, evidence for broad cognitive enhancement or dementia prevention is weak, with most studies suffering from small samples, inadequate controls, and limited follow-up periods. The distinction between practiced task improvement and genuine neural protection is crucial and largely unresolved.

Evidence Quality

3

Meta-Analyses

15

RCTs

5

Observational

Important Caveats

  • โš ๏ธ Most improvements are confined to trained tasks with minimal transfer to daily function
  • โš ๏ธ The ACTIVE dementia finding was a secondary analysis of one training type
  • โš ๏ธ Commercial brain training claims significantly overstate the evidence
  • โš ๏ธ Long-term neuroprotective effects remain largely unproven in well-controlled studies
  • โš ๏ธ Comparison with simpler activities like reading or socializing is rarely made

Population Studied

Primarily healthy older adults aged 60-85; some studies include those with mild cognitive impairment (MCI); limited data in younger populations

Dosage

Typically 2-5 sessions per week, 30-60 minutes per session; ACTIVE trial used 10 sessions over 5-6 weeks with optional booster training at 11 and 35 months

Duration

Training programs range from 4-12 weeks; follow-up in ACTIVE trial extended to 10 years; most studies have less than 1 year follow-up

Supporting Studies (3)

Computerized Cognitive Training in Cognitively Healthy Older Adults: A Systematic Review and Meta-Analysis of Effect Modifiers

Meta-Analysis

Lampit A, Hallock H, Valenzuela M. ยท PLoS Medicine (2014)

Meta-analysis of 52 RCTs (4,885 healthy older adults) found that computerized cognitive training produced small but significant improvements (g = 0.22) in overall cognitive function, with larger effects for training frequency of 3+ sessions/week.

View paper (DOI) โ†’

Ten-Year Effects of the Advanced Cognitive Training for Independent and Vital Elderly Cognitive Training Trial on Cognition and Everyday Functioning in Older Adults

RCT

Rebok GW, Ball K, Guey LT, et al. ยท Journal of the American Geriatrics Society (2014)

Ten-year follow-up of the ACTIVE trial (2,832 participants) showed that each training group maintained significantly better performance in its targeted cognitive domain compared to controls, with self-reported improvements in daily function.

View paper (DOI) โ†’

Speed of processing training results in lower risk of dementia

RCT

Edwards JD, Xu H, Clark DO, et al. ยท Alzheimer's & Dementia: Translational Research & Clinical Interventions (2017)

Secondary analysis of ACTIVE trial data found that speed-of-processing training was associated with 29% lower risk of incident dementia after 10 years compared to control (HR 0.71), though memory and reasoning training showed no such benefit.

View paper (DOI) โ†’

Contradicting Studies (1)

Do 'Brain-Training' Programs Work?

Meta-Analysis

Simons DJ, Boot WR, Charness N, et al. ยท Psychological Science in the Public Interest (2016)

Comprehensive review concluded that evidence does not support claims that brain training programs prevent cognitive decline or dementia. While training improves performance on practiced tasks, transfer to broad cognitive function or real-world outcomes is not reliably demonstrated.

Why this disagrees:

This consensus paper argues that improvement on trained tasks should not be confused with genuine neuroprotection or cognitive enhancement. Methodological issues including inadequate controls, placebo effects, and limited transfer evidence undermine strong claims about dementia prevention.

View paper (DOI) โ†’
Share:
Was this helpful?

Related Claims