Brain training sped up walking — but only when multitasking

Physical and cognitive decline tend to travel together in later life, and the two are closely interrelated. Most research so far has asked the question one way round: does physical exercise sharpen the brain? Far fewer studies have asked whether training the brain changes how the body moves. That matters because walking is not purely automatic — it draws on attention and executive control, particularly when someone is also talking, planning a route or scanning traffic. If strictly non-physical practice could make walking more robust, it would be relevant to older adults who cannot easily exercise.

The reviewers searched PubMed, Embase, CINAHL and PsycInfo in February 2025 for randomised controlled trials of what they define as pure cognitive training: strictly non-physical, guided practice on a standardised set of cognitive tasks. Fourteen trials in older adults met the criteria. Outcomes covered standing balance tests and comprehensive balance scales, gait speed while walking only, gait speed while performing a cognitive task at the same time, and the Timed Up and Go test. Methodological quality was scored with the PEDro scale and certainty rated with GRADE; meta-analysis was restricted to higher-quality studies where at least three were available.

The primary signal was narrow. Cognitive-motor dual-task gait speed improved significantly, with a standardised mean difference of 0.39 (95% CI 0.04-0.75) across 376 participants, rated moderate-quality evidence. Single-task gait speed did not improve (SMD 0.12, 95% CI -0.05 to 0.30; 506 participants, moderate quality). Functional mobility showed no significant gain (SMD 0.29, 95% CI -0.60 to 1.18; 566 participants, low quality), nor did balance (SMD 0.18, 95% CI -0.34 to 0.69; 139 participants, very low quality). Eight of the 14 studies (57%) were rated good quality and 6 (43%) fair. The authors noted that protocols targeting executive function, with 40- to 60-minute sessions 2 to 3 times weekly over 8 to 10 weeks, were the ones shown to work.

This runs the familiar exercise-and-cognition question in reverse and finds transfer, but only where the task itself is cognitively loaded. That pattern fits the view that walking while thinking is limited by attentional capacity rather than by the legs, and it does not suggest cognitive training builds any underlying physical capacity. It complicates the marketing story around brain training by showing where the benefit stops. Open questions remain over whether the dual-task gain reduces real-world trips and falls, how long it lasts after training ends, and whether the very low-certainty balance estimate would change with larger trials.

The reasonable reading is that mental training appears to transfer to a specific skill — walking while your attention is occupied — rather than to movement in general. It is not a substitute for physical training if the goal is balance, mobility or plain walking speed, because those outcomes did not shift. The certainty behind the balance and mobility estimates was low to very low, so those nulls are provisional rather than final. For anyone interested in ageing well, this is a reason to think the cognitive side of walking is trainable, not a reason to swap exercise for puzzles.