Four-Year Trial Ties Structured Exercise to Better Cognitive Scores in Healthy Seniors, Weaker Signal in Those With Memory Complaints

A single-blind randomized trial that followed 882 older adults in Italy for 48 months found higher Mini-Mental State Examination scores among cognitively healthy participants assigned to supervised exercise, but a smaller and statistically borderline effect in those with subjective memory complaints, and too few participants with mild cognitive impairment to draw conclusions.

A randomized controlled trial that tracked participants for four years, one of the longest dedicated exercise-only prevention trials published to date, found that cognitively healthy older adults assigned to a supervised exercise program scored significantly higher on a standard cognitive test after 48 months than those who kept their usual habits. The results, from the Physical Exercise for Prevention of Dementia (EPD) study, appeared in Frontiers in Aging on August 10, 2026, and are registered with ClinicalTrials.gov as NCT02236416.

The trial enrolled 882 community-dwelling adults aged 50 and older in Campobasso, Italy, sorted at entry into three groups: cognitively healthy, those reporting subjective memory complaints (SMC), and those with mild cognitive impairment (MCI). Within each group, participants were randomized 1:1 to 48 months of supervised, moderate-to-high-intensity exercise combining aerobic work, resistance training and stretching, or to a control arm that made no prescribed change to activity. Of the 882 randomized, 674 completed the full four years and were included in the final analysis. The primary measure was change in Mini-Mental State Examination (MMSE) score, a standard 30-point screening test for cognitive function, tracked alongside eight other neuropsychological tests at 12, 24, 36 and 48 months.

Among healthy participants, the exercise group showed significantly higher MMSE scores than controls (between-group effect p equals 0.008; time-by-group interaction p less than 0.001), with similar advantages on secondary tests of memory, attention and reasoning. Among those with subjective memory complaints, exercisers also scored higher on the MMSE, but the researchers reported that difference "only approached statistical significance" (p equals 0.052). The mild cognitive impairment subgroup, the group closest to a dementia diagnosis, was too small to allow the authors to draw firm conclusions, a limitation they attribute partly to a relatively strict MMSE entry cutoff that may have screened out eligible volunteers.

The study's authors, based at the University of Molise, call the healthy and SMC findings evidence that a long-term, structured exercise program can help slow age-related cognitive decline, and say a larger, adequately powered trial in people with mild cognitive impairment is still needed before anything can be said about that group.