Dual-Task Exercise May Sharpen the Aging Brain More

Walking is good for older adults. Walking while reading a map, choosing a route, and correcting a wrong turn may do more for executive function than walking the same terrain without those decisions. In a 24-week randomized trial, older adults assigned to orienteering improved executive function and visuospatial skills by 49.2%, versus 16.5% in a hiking group; the group-by-time difference was statistically significant (p<0.001). Hiking helped, too. The point is the size of the added cognitive gain, not that an ordinary hike somehow fails the brain.

That result fits a wider, still imperfect literature: exercise that makes people move and think at the same time often improves executive function and dual-task walking more than movement alone. It does not establish that every brain-training add-on improves every mobility outcome. The largest trial in this area found a selective improvement in response inhibition but no added benefit for gait speed, walking endurance, or dual-task gait cost.

Same trail, different cognitive work

Biehl-Printes et al. (2023) randomized 90 community-dwelling older adults in Brazil to 24 weeks of orienteering or hiking. The comparison matters. Both groups walked outdoors on natural terrain for the same session duration. Orienteering added ongoing map-based navigation; hiking did not.

Both groups improved, but the orienteering group improved more on several cognitive measures. Executive function and visuospatial skills rose 49.2% with orienteering versus 16.5% with hiking (effect size 2.00; p<0.001). Processing speed rose 31.4% versus 14.7% (effect size 1.64; p<0.001), and memory rose 39.8% versus 6.5% (effect size 1.62; p<0.001).

The study also tested mobility under cognitive load. Orienteering reduced timed up-and-go dual-task performance by 6 seconds, a 42% improvement, and had a between-group effect size of 0.56 for dual-task cost (p=0.001). Dual-task versions that counted cognitive stops and errors produced between-group effect sizes of 0.85 and 0.90. These outcomes fit everyday mobility better than a quiet, straight-line walk: people cross streets, find places, talk, and notice hazards while moving.

The effect sizes for cognition were unusually large for exercise research. That should make readers interested, not credulous. The participants were sedentary at baseline, the sample was modest, and the study took place in one country. Time in nature and social interaction were not fully matched between groups. A low-baseline population can have more room to improve, and a sensitive testing battery can make change look larger. The trial supports a meaningful advantage for this specific navigation-based program; it does not give a universal percentage gain to expect from downloading a trail app.

Why integrated challenge may matter

Navigation, reactive sports, and movement games require attention, spatial decisions, error correction, and inhibition while the body is already busy. That differs from doing a workout and later opening a cognitive-training app. St George et al. (2022) found that inhibition tasks produced more interference with dual-task walking and larger increases in prefrontal cortical activity than working-memory tasks in young and older adults. Selecting a route while suppressing the tempting but wrong turn plausibly places this kind of demand on executive control, although that link to orienteering remains an inference rather than a direct measurement.

A smaller randomized trial offers relevant brain-imaging evidence. In Liao et al. (2025), 39 older adults with cognitive frailty completed 12 weeks and 36 sessions of either interactive boxing-cycling (20 participants) or stationary cycling (19 participants). The interactive program combined cycling with visual target tracking. Compared with cycling alone, it produced better dual-task walking speed, lower walking cost, and better domain-specific cognitive function. Both groups improved on general frailty measures.

The boxing-cycling group also showed lower prefrontal oxygenated hemoglobin on functional near-infrared spectroscopy after training. Because performance improved while oxygenated hemoglobin fell, the authors interpreted the pattern as greater neural efficiency: less prefrontal effort for the task. That is a reasonable interpretation, not proof that the oxygenation change caused the behavioral gains. Salzman et al. (2023) describes why context matters: higher prefrontal activation can reflect compensation in some older adults, while lower activation can indicate efficiency only when performance is maintained or improves.

The Liao trial is especially relevant to people with cognitive frailty, not automatically to cognitively healthy adults. Its 39-person sample also cannot settle who benefits most. Still, it supports the idea that a cognitively embedded movement task can change both performance and a prefrontal biomarker more than predictable cycling.

The larger trial draws a hard boundary

The 12-month PASSWORD randomized trial provides the most useful check on an overly broad dual-task story. It enrolled 314 Finnish older adults with a mean age of 74.5 years; 93% completed the study. Everyone received comprehensive physical training: supervised walking and balance sessions, resistance training, home exercise, and 150 minutes a week of aerobic activity. One group also received computer-based executive-function training for 15 to 20 minutes, prescribed three to four times weekly.

Adding the computer training improved Stroop response inhibition by 6.9 seconds relative to physical training alone (95% CI, -13.0 to -0.8). It did not significantly improve gait speed: the between-group difference was 0.02 m/s (95% CI, -0.03 to 0.08). It also did not significantly improve six-minute walking distance (-3.8 m; 95% CI, -16.9 to 9.3) or dual-task gait cost (-0.22; 95% CI, -1.74 to 1.30). Task switching on Trail Making B-A showed no added benefit.

This was not a failure of physical training. It was a failure of added, separately delivered computer training to transfer broadly to locomotion. Adherence to the executive-function program averaged about 1.9 sessions a week rather than the prescribed three to four, which may have weakened its effect. But the intent-to-treat null results remain the strongest available evidence against promising faster walking simply by bolting brain exercises onto an exercise routine.

The contrast with orienteering and boxing-cycling may reflect delivery rather than contradiction. Those interventions embed cognitive decisions inside movement; PASSWORD delivered physical and cognitive training as distinct components. The populations, endpoints, and programs also differed. No trial cited here directly compares physically integrated cognitive challenge with the same dose of separately delivered cognitive training while measuring both gait and brain activity.

What the broader literature supports—and what it cannot yet answer

A 2026 scoping review by Vasu et al. examined 31 studies published from 2014 through 2024 in community-dwelling adults aged 60 and older. It found that motor-cognitive dual-task interventions generally outperformed single-task interventions for balance, gait, and executive function. The review used PRISMA-ScR methods and rated study quality with the Mixed Methods Appraisal Tool, but 9 of 31 studies received low or very low ratings. Different populations, interventions, and outcome measures prevented a formal pooled dose-response estimate. "Do more dual-task training" is supported more strongly than any exact weekly prescription.

Other evidence also limits a simple prefrontal explanation. Rosso et al. (2024) reported that physical therapy improved gait speed in a randomized trial without changing prefrontal activation during walking. Gait speed can improve through physical and motor pathways without detectable prefrontal change. And Rosso et al. (2025) found that nigrostriatal dopaminergic integrity was independently related to prefrontal activity and dual-task gait performance in older adults. That biology may affect how much prefrontal compensation a person needs and how readily it changes; exercise does not directly repair dopaminergic integrity.

Finally, simultaneous training may not always beat sequential multicomponent training. In a 42-person randomized trial of people with mild cognitive impairment, Bilgin et al. (2022) found both approaches improved cognition, processing speed, gait speed, and balance, without consistent superiority of multitask training across outcomes.

A practical way to use the evidence

If your goal is general fitness, steady cycling, hiking, resistance training, and walking remain worthwhile. Physical activity improves physical health and can support cognition. If your goal includes executive function or staying steady when life adds distractions, choose some sessions that force real-time decisions during movement.

• Navigate an unfamiliar walking route with a map rather than following a memorized loop.
• Try dance, racquet sports, interactive cycling, or agility drills that require responses to changing cues.
• Scale difficulty to safety. A confusing route is training; a fall risk is poor program design.
• Do not expect a cognitive app added after exercise to reliably increase gait speed.

The evidence favors movement that makes the brain solve problems in real time. It does not support treating every added mental task as a shortcut to faster walking or a younger brain.

The most defensible conclusion is modest but useful: for many older adults, exercise with an integrated cognitive challenge can produce larger gains in executive function and dual-task mobility than comparable single-task exercise. The best next study would test that proposition at scale, with standardized programs, long-term follow-up, and outcomes that distinguish better test performance from safer, more independent daily life.