A combined approach of regular exercise, healthy eating, staying cognitively and socially active, not smoking, and treatment of high blood pressure, diabetes, and high cholesterol most demonstrably reduces the risk of dementia. Single measures have little effect; a structured multidomain programme works significantly better than self-management alone.
Dementia cannot be prevented entirely, but research shows that a large proportion of cases are linked to modifiable risk factors. Roughly one third of all dementia cases is associated with factors you can influence yourself. That makes prevention worthwhile, even if it offers no guarantee (PMID 32738937, 31964800, 32780691).
The strongest individual modifiable risk factors are: physical inactivity, smoking, untreated high blood pressure in middle age, type 2 diabetes, and high cholesterol. Each of these factors has been linked to a considerably higher dementia risk in large-scale research. High blood pressure and diabetes raise the risk by almost two to more than two times. Smoking is associated with an almost two-fold higher dementia risk and a 27% higher risk of mild memory problems. These are therefore exactly the conditions your GP would already be treating you for (PMID 33271079, 32738937, 30291317).
Beyond physical health, social and mental factors also matter. Living in social isolation, such as living alone or becoming widowed early, is associated with more than twice the dementia risk. Depression is likewise a modifiable risk factor, although it has not yet been proven that treating depression directly lowers dementia risk. Low levels of education are strongly associated with higher dementia risk; staying cognitively active throughout life appears to be protective (PMID 33271079, 32738937, 30291317).
Regarding exercise: both aerobic training (walking, cycling, swimming) and strength exercises show protective brain mechanisms in research, including improved cerebral blood flow, clearance of harmful proteins, and production of growth factors for brain cells. The evidence is encouraging but not yet conclusive for definitive dementia prevention. Physical inactivity is considered one of the best-supported modifiable risk factors (PMID 32780691, 31964800).
Most studies testing just one lifestyle measure at a time, such as adjusting diet alone or increasing exercise alone, produced disappointing or modest results. Two large randomised trials, the Finnish FINGER study and the American US POINTER study (2025, n=2111), show that a combined approach, comprising guided exercise, a healthy diet, cognitive training, social activities, and monitoring of blood pressure and cholesterol, does produce measurably better outcomes than an unstructured approach. The benefit in US POINTER was statistically significant and greatest in people who already had somewhat more difficulty with thinking and memory. Two other large European studies (MAPT, PreDIVA) found no overall significant effect but did find positive results in higher-risk subgroups (PMID 40720610, 30291317).
The claims are based on several large observational cohort studies (including a Chinese cohort with n=46,011) and randomised controlled trials (FINGER, US POINTER, MAPT, PreDIVA). The evidence is predominantly moderately strong: associations are consistent but largely observational; causality for most individual risk factors has not been definitively proven. The multidomain intervention studies provide the strongest evidence for a causal relationship.