Investigators reported that slower quick walking speed (QWS) was associated with higher hazard of incident dementia over a 10-year follow-up, while usual walking speed (UWS) did not reach significance for dementia in this cohort. The dementia-free sample comprised 1,148 participants aged 60 years or older (mean age 68.5 years; 53% women). Per-standard-deviation slower QWS, the hazard ratio for incident dementia was 1.34 (95% CI 1.10–1.63), a result the authors described as statistically significant. UWS produced a hazard ratio of 1.20 (95% CI 0.98–1.46), which crossed 1.0 and therefore was not significant in this analysis. The study team suggested that a brief quick walking trial might reveal risk that habitual-pace screening overlooks.
The cohort analysis linked both slower QWS and slower UWS with poorer executive function and with higher odds of elevated depressive symptoms. For the binary outcome of high depressive symptoms, QWS carried an odds ratio of 1.24 (95% CI 1.07–1.45) and UWS an odds ratio of 1.26 (95% CI 1.09–1.46). The study used linear regression for continuous cognitive outcomes and logistic regression for depressive symptom thresholds; investigators reported associations per SD decrement in walking speed. The authors also noted that none of the gait measures—QWS, UWS, or adaptability—showed an association with delayed verbal recall in this sample, indicating a domain-specific relationship concentrated on executive function rather than memory recall.
Data indicated that both slower QWS and UWS were associated with greater hazard of incident stroke across the full analytic sample (n = 2,308; mean age 61.5 years; 54% women). Reported hazard ratios were QWS HR 1.30 (95% CI 1.03–1.64) and UWS HR 1.36 (95% CI 1.11–1.65). The investigators observed that stroke and executive function associations were driven primarily by outcomes in men; the stroke association for QWS in men was reported as HR 1.55 (95% CI 1.11–2.17). The paper therefore describes sex-specific heterogeneity in some cerebrovascular associations, with stronger relationships seen in men for incident stroke in this dataset.
Gait adaptability—defined by the authors as the difference between QWS and UWS—was associated with worse executive function overall and within sex-stratified analyses for women and men. For incident dementia, lower gait adaptability was associated with increased hazard in women (HR 1.35, 95% CI 1.04–1.76), whereas it was not associated with depressive symptoms or with incident stroke according to the reported results. The investigators emphasized that adaptability captured a different signal than absolute walking speed measures, suggesting that the capacity to increase pace above usual walking speed may reflect specific neurological reserve or vulnerability relevant to cognitive outcomes in women.
The authors examined participants from the Framingham Heart Study Offspring cohort using two overlapping analytic samples. A larger sample (n = 2,308) supported cross-sectional cognitive analyses and incident stroke modeling. A dementia-free subset of 1,148 participants aged 60 years and older was used for incident dementia analyses. Mean ages, standard deviations, and sex distributions were reported for each analytic group. Associations were estimated per SD decrement in walking speed or adaptability using linear models for continuous cognitive measures, logistic regression for elevated depressive symptoms, and Cox proportional hazards models for 10-year incident stroke and dementia outcomes. The study team disclosed funding sources and ethical approval from the Institutional Review Board at Boston Medical Center. Authors also stated that the manuscript is a preprint and has not undergone peer review.
The manuscript is a preprint and has not been certified by peer review; the authors explicitly caution against using these results to guide clinical practice. The report provides effect estimates and confidence intervals but does not present additional mechanistic or interventional data in this version. The source did not report details on gait trial protocol specifics beyond the definitions of UWS, QWS, and gait adaptability in the abstract; calibrated thresholds, exact trial distances, and reproducibility metrics were not described in the abstract and would need to be referenced in the full text or supplementary materials. Open questions remaining in this report include whether quick-walk assessment improves prediction beyond established clinical risk scores and how these associations perform across diverse populations. The authors noted data are available on request and listed funders and competing interest statements; further peer review and external validation were not reported in the preprint version.