Walking, Dementia, and HFpEF: What Step Data Show

Walking more is associated with lower dementia risk, and the association has a useful shape: risk fell sharply as daily steps rose, then leveled off near 9,826 steps per day. In the same broad category of behavior, more movement and less sitting were associated with lower risk of heart failure with preserved ejection fraction, or HFpEF, in older women. These findings do not prove that a step counter can prevent dementia or heart failure. They do offer more practical detail than the usual instruction to "exercise more."

The important detail is intensity. A deliberate brisk walk appears more strongly associated with lower dementia risk than the same number of casual steps. For HFpEF, even light activity was associated with lower risk. The two results point toward a sensible routine: accumulate steps, include a brisk stretch if you can, and spend less of the day sitting.

Dementia risk fell until roughly 9,826 daily steps

In a prospective UK Biobank study of 78,430 adults, wrist accelerometers measured activity for about seven days. Researchers then linked the data to hospital and death records over a median 6.9 years. Higher daily step counts were associated with lower incidence of all-cause dementia in a nonlinear dose-response pattern.

In the 2022 JAMA Neurology study by Del Pozo Cruz et al., the modeled optimum was about 9,826 total steps per day. At that point, the hazard ratio for incident dementia was 0.49 compared with the low-activity reference group (95% CI, 0.39–0.62). The curve plateaued around that level, so the data did not show an additional benefit above approximately 9,826 steps.

A lower benchmark may matter more for people starting from a sedentary baseline. About 3,826 steps per day corresponded to a hazard ratio of 0.75 (95% CI, 0.67–0.83), which the authors estimated captured roughly 49% of the maximum modeled risk reduction. That is a better interpretation of the results than treating 10,000 steps as a biological commandment carved into a treadmill. The familiar target is close to the modeled plateau, but it is not a threshold where benefits suddenly begin.

Brisk steps carried the larger association

Total volume was not the whole story. The researchers separated purposeful steps, defined as at least 40 steps per minute, from incidental steps below that cadence. At their respective modeled optima, purposeful steps showed a hazard ratio of 0.43 (95% CI, 0.32–0.58) at about 6,315 daily steps, while incidental steps showed a hazard ratio of about 0.58. Both categories were associated with lower dementia risk, but the association was stronger for the faster steps.

Peak 30-minute cadence produced the steepest curve. A cadence near 112 steps per minute was associated with a hazard ratio of 0.38 for dementia (95% CI, 0.24–0.60). That pace is not a sprint. In the CADENCE-adults laboratory study, 100 steps per minute corresponded to moderate intensity, at least 3 METs, while 130 steps per minute corresponded to vigorous intensity. A 112-step-per-minute stretch therefore falls around the transition from moderate toward vigorous walking.

The result does not mean everyone needs exactly 30 continuous minutes at 112 steps per minute. The study examined each participant's highest-cadence 30 minutes, not a prescribed workout. It does suggest that a daily brisk-walking segment may matter more than simply padding a step count with slow laps around the kitchen.

Heart failure is not one outcome

Heart failure with preserved ejection fraction and heart failure with reduced ejection fraction have different biology and should not be treated as interchangeable endpoints. HFpEF often involves impaired relaxation and filling of the heart; HFrEF involves impaired pumping function, often after ischemic myocardial injury. That distinction showed up in accelerometer data from older women.

In a 2024 OPACH prospective cohort study, 5,951 women aged 63 to 99 wore accelerometers for seven days and were followed for a mean of 7.5 years. Physicians adjudicated heart-failure events by subtype. Each standard-deviation increase in daily steps was associated with lower overall heart-failure risk (HR, 0.74; 95% CI, 0.63–0.88) and lower HFpEF risk (HR, 0.71; 95% CI, 0.57–0.88).

Sitting showed the opposite pattern for HFpEF. Each standard-deviation increase in total sedentary time was associated with a 29% higher HFpEF risk (HR, 1.29; 95% CI, 1.10–1.51). Light-intensity physical activity was also associated with lower HFpEF risk (HR, 0.80 per standard-deviation increase; 95% CI, 0.70–0.93). The associations did not materially differ by age, race or ethnicity, body mass index, physical function, or comorbidity score within this cohort.

The HFrEF result was different: daily steps had a hazard ratio of 0.83, but its 95% CI ran from 0.62 to 1.12. That result was not statistically significant. It does not establish that walking fails to help prevent HFrEF; only 110 adjudicated HFrEF cases occurred, leaving wide uncertainty around the estimate.

The subtype pattern is biologically plausible but remains an association, not proof of mechanism. Research on HFpEF links exercise intolerance to left-atrial stiffness in older adults with obesity, as described in a 2022 study of left-atrial stiffness. A mechanistic review of HFpEF also describes obesity, diabetes, visceral fat, inflammation, fibrosis, and myocardial stiffness as relevant pathways. Habitual movement could plausibly affect some of those cardiometabolic factors, but neither source shows that a particular daily step count changes them or prevents HFpEF.

Light movement still counts

The OPACH findings should reassure people who cannot sustain fast walking. A 2022 meta-analysis of objectively measured light-intensity activity found an inverse association with cardiovascular mortality. That evidence does not provide a discrete HFpEF step target, because OPACH reported effects per standard-deviation increase rather than a threshold or plateau. Still, it supports the practical conclusion that replacing some sitting with light movement is worthwhile, especially for older adults.

Evidence beyond these cohorts also supports physical activity for cognitive health in general. A 2024 systematic review and meta-analysis found that long-term exercise and multidomain interventions were associated with lower risk of mild cognitive impairment and dementia in older adults. Its varied exercise programs, populations, and durations cannot validate the specific 3,826- and 9,826-step estimates from UK Biobank. They do make the broader relationship less dependent on one study.

What these data can and cannot tell you

Both central studies were observational. People who walk more may differ from less-active people in ways statistical adjustment cannot fully remove. The dementia study excluded diagnoses occurring within two years and adjusted for several cardiometabolic factors without materially changing its findings. Yet preclinical dementia can reduce activity five to 10 years before diagnosis, so reverse causation remains plausible.

The dementia cohort was 96.7% White, and the exact step thresholds have not been replicated in more diverse or lower-income populations. The HFpEF study included a racially and ethnically diverse group of women, but it did not include men. Both studies measured activity during one roughly week-long window, years before outcomes, which may misclassify long-term habits.

Cadence also deserves humility. The purposeful-step cutoff of 40 steps per minute is a population-average classifier, not a personalized physiological test. Two 2024 analyses comparing cadence-based and machine-learning classifications found only partial agreement and different associations with mortality, including in UK Biobank data: one analysis and a related UK Biobank analysis. A fixed cadence can therefore describe a useful population pattern without perfectly capturing intensity for every older adult or gait style.

For a practical, evidence-calibrated goal: build toward roughly 9,800 daily steps if feasible, include a brisk walking period that approaches 112 steps per minute, and replace some prolonged sitting with movement. Treat those numbers as observational benchmarks, not guarantees or medical treatment.

If you have heart disease, exertional symptoms, mobility limits, or a substantial fall risk, ask a clinician how to increase activity safely. For everyone else, the data support a less glamorous but more defensible intervention than most longevity products: walk regularly enough that your watch has something real to report.