Age is just a number: Clustering gait and functional measures.

Abstract

Objective: As we age, we walk slower, but it remains unclear whether this is consistent at an individual level. Current clinical assessment of function assumes movement deficits with older age, and clinical norms are linked to decades or specific age stratifications such as "old" or "oldest-old". Current approaches stratifying by age may hide trends of higher and lower functioning individuals within each age bracket. Therefore, our aim was to cluster spatiotemporal data, from the 1000 Norms Project, to understand if patterns of function could be identified without using age as a factor.

Methods: The 1000 Norms Project, a cross-sectional, observational study, collected gait, functional performance, and self-reported health data (participants (n = 695) aged 18-92 years). Spatiotemporal and functional data were clustered, after rendering the parameters dimensionless.

Results: Three clusters were identified (n = 277, 208, 210). Although age significantly differed between clusters, each showed a broad range (e.g. 20-92 years). Additionally, walking speed (Froude number) did not differ between clusters, often used to separate by age. Our clusters defined 3 groups, 'higher functioning', 'age average' and 'cautious gait', whose spatiotemporal, functional performance, strength and quality of life measures vastly differed, independent of walking speed and including a wide range of ages.

Conclusion: Our analysis suggests that age should not be used to separate individuals into groups, and that our assumption of "age matters" may not be relevant when determining true functional movement ability. Further work is needed to understand normal senescence, true negative loss, and reversible loss within these functionally different groups.