3,800 steps linked to 50% dementia benefit

HuffPost UK reported on August 6 that a 2022 JAMA Neurology study linked relatively modest daily walking totals with lower dementia risk, reviving attention on a threshold far below 10,000 steps. The underlying paper followed 78,430 adults in the UK Biobank and found that dementia risk fell as daily step counts rose, with benefits leveling off at about 9,800 steps a day. The same study identified 3,826 steps a day as the point at which participants reached half of the maximum observed reduction in risk. The findings were observational, not a clinical trial, and the authors did not say walking itself proved cause and effect. ### Where does the 3,800-step figure come from? A September 6, 2022, paper in JAMA Neurology by Borja del Pozo Cruz, Matthew Ahmadi, Sharon L. Naismith and Emmanuel Stamatakis examined accelerometer-measured step counts and later dementia diagnoses in UK adults aged 40 to 79. The researchers used wrist-worn activity trackers and linked those data to registry-based dementia outcomes through October 2021. (jamanetwork.com) The paper reported a “minimal dose” of 3,826 steps per day, defined as the point at which dementia risk had reached half of the maximum reduction seen in the study. MedPage Today, summarizing the paper at the time, reported that level corresponded to a hazard ratio of 0.75, while the “optimal dose” was 9,826 steps per day with a hazard ratio of 0.49. ### Did the study say 3,800 steps cuts dementia risk by 50%? (jamanetwork.com) The JAMA Neurology paper did not say 3,800 steps cut dementia risk in half. It said 3,826 steps represented half of the maximum observed risk reduction in the cohort, compared with the larger reduction seen near 9,826 steps a day. Harvard Health, describing the same study, said people who walked about 9,800 steps per day were 51% less likely to develop dementia than people with very low step counts, while people who walked about 3,800 steps per day were about 25% less likely to develop dementia. (medpagetoday.com) That is why some later coverage describes 3,800 steps as delivering “50% of the benefit,” rather than a 50% drop in dementia risk on its own. (jamanetwork.com) ### How strong was the evidence? The study included 78,430 adults and had a median follow-up of 6.9 years, which makes it large for this kind of activity research. The authors adjusted for age, sex, education, smoking, alcohol use, diet, medication use, sleep and other factors, and they excluded dementia diagnoses in the first two years in a main analysis to reduce the chance that early disease had already reduced participants’ activity. (health.harvard.edu) JAMA Neurology and Europe PMC both identify the work as a prospective cohort study, which means it can show an association but cannot prove that increasing steps alone prevents dementia. The researchers wrote that more daily steps were associated with “steady declines” in dementia incidence risk up to about 9,800 steps per day, after which benefits plateaued. ### What does this mean for the 10,000-step target? (jamanetwork.com) The 9,826-step figure in the paper sits close to the popular 10,000-step benchmark, but the study’s lower threshold drew attention because it suggested measurable benefit started much earlier. HuffPost UK’s August 6 piece used that finding to argue that people do not need to hit 10,000 steps before brain-health benefits begin. (jamanetwork.com) The study also looked at intensity, and MedPage Today reported that “purposeful steps” — roughly a faster cadence — were associated with lower dementia risk as well. The authors said step-based recommendations may be appropriate for dementia-prevention guidelines, though the paper itself stopped short of prescribing a single target for everyone. ### What should readers watch next in the research? (jamanetwork.com) The 2022 JAMA Neurology paper remains the source for the 3,826-step figure, and later coverage has largely been a restatement of that result rather than new trial data. Readers looking for the exact methods, hazard ratios and participant definitions will find them in the original JAMA Neurology report by del Pozo Cruz and colleagues, published online on September 6, 2022. (jamanetwork.com) (medpagetoday.com)