Exercise and the brain: what training can do for your mind — and what nobody may promise

The contradiction on which this topic turns: observational studies say yes, the big randomised trial says no. In cohorts — most recently a meta-analysis of 49 studies with 2.86 million people (2026) — physically active people receive dementia diagnoses noticeably less often, and long sitting time is linked to higher risk. Sounds clear. But observation cannot separate cause from consequence — and that is exactly the catch: dementia begins in the brain many years before diagnosis, and one of its earliest unnoticed consequences is that people become less active. A 2023 analysis (of Alzheimer’s disease, the most common form of dementia) shows the pattern clearly: the protective association is large over short follow-up and disappears when people are followed for 15 years or more — the further the diagnosis sits from the activity measurement, the smaller the "protection". Our reading, shared by many in the field: that smells like reverse causation.

The hardest test came from the LIFE trial (JAMA 2015): 1,635 previously sedentary people aged 70 to 89 were randomly assigned to a two-year structured exercise programme or a health-education programme. Result: no difference in cognitive tests, no difference in new cases of mild cognitive impairment or dementia (13.2 versus 12.1 percent). One trial does not settle a research question — two years may be too short for a disease with decades of lead time, and the control group was not idle. But it is the best randomised evidence in existence, and it forbids the promise the fitness industry most likes to make.

The good news needs no exaggeration. First: training improves thinking performance in the present. Randomised trials in cognitively healthy older adults show small but real improvements in executive functions — the ability to plan, switch and concentrate. Second, the unspectacular, solid route: training demonstrably improves several of the risk building blocks the Lancet Commission names for dementia — blood pressure (the vascular factor for the brain), excess weight, and through classes and clubs social participation too. Third, strength and balance training prevents falls — and head injuries are themselves a dementia risk factor.

So the honest summary is not "training does nothing for your brain" — it is: training demonstrably works on the brain's surroundings; on the disease itself it is unproven. That is a strong argument for keeping at it. It is just not a sales pitch for miracles.

The 2020 Lancet Commission systematically compiled the modifiable dementia risk factors: early-life education, hearing loss, high blood pressure, obesity, smoking, depression, social isolation, physical inactivity, diabetes, alcohol, head injury, air pollution. Around 40 percent of dementia cases worldwide are statistically associated with these twelve factors combined — a population-level model, not an individual prognosis. Physical inactivity is one of twelve — and not the largest (in this model that is midlife hearing loss, followed by education).

Practically, that means: anyone who wants to do something for their brain has a list, and training is on it — alongside wearing hearing aids when needed, treating blood pressure, stopping smoking, social contact and treating depression. The WHO guideline on risk reduction (2019) recommends exactly this bundle. No single building block is the miracle — the bundle is the strategy.

The most important limit of this article: with existing memory or orientation problems, training is not an answer to the question "what is going on?" New or increasing forgetfulness that disrupts daily life, word-finding or orientation problems, personality changes — that belongs promptly in GP and, if needed, neurological assessment, not least because some causes (thyroid, vitamin B12 deficiency, depression, medication side effects) are treatable. Exercise remains valuable after a diagnosis too — for function, mood and independence — but then as part of a medically guided overall plan, not as a substitute for one.

And the closing formula that runs through this magazine: move — for strength, bones, blood pressure, falls, sleep, and because you feel better today. Those are proven reasons enough. You do not need a dementia promise for that — and anyone who makes you one has not read the evidence or is ignoring it.