Background: Gait impairments are associated with falls and loss of independence. The study of factors associated with poorer gait may assist in developing methods to preserve mobility in older people. The aim of this study was to examine the associations between a range of cognitive functions and gait and gait variability in a population-based sample of older people.

Methods: Gait and intra-individual gait variability measures were obtained using the GAITRite walkway in a sample of older people, aged 60–85 years (N = 422), randomly selected from the Tasmanian electoral roll. Raw scores from a cognitive battery were subjected to principal component analyses deriving four summary domains: executive function/attention, processing speed, memory, and visuospatial ability. Multivariable linear regression was used to examine associations between cognitive domains and gait measures adjusting for age, sex, ambulatory activity, medication use, and mood.

Results: The mean age of the sample was 72.0 years (SD = 7.0), with 238 men (56%). Poorer executive function was independently associated with poorer performance in most absolute gait measures and with greater variability in double support phase and step time. Processing speed was associated with absolute gait measures and double support phase variability. Visuospatial ability was only associated with greater double support phase variability, independently of executive function and processing speed. Memory was not independently associated with any gait measure.

Conclusions: In community-dwelling older people, executive function/attention and processing speed were associated with many aspects of gait, whereas visuospatial ability may only play a role in double support phase variability.