Laboratory-based gait variability and habitual gait entropy do not differentiate community-dwelling older adults from those with subjective memory complaints

Background: Age-related cognitive decline may be delayed with appropriate interventions if those at high risk can be identified prior to clinical symptoms arising. Gait variability assessment has emerged as a promising candidate prognostic indicator, however, it remains unclear how sensitive gait variability is to early changes in cognitive abilities. Research question: Do community-dwelling adults over 65 years of age with subjective memory complaints differ from those with no subjective memory concerns in terms of laboratory-measured or free-living gait variability? Methods: This cross-sectional study recruited 24 (age = 73.5(SD 6.4) years) community-dwelling people with subjective memory complaints and twenty seven (age = 70.9(4.3) years) individuals with no subjective memory concerns. A sample of 9 individuals with diagnosed mild dementia were also assessed (age = 86.5(7.0) years). Gait variability was assessed in a laboratory during walking at preferred pace (single-task) and while counting backward by seven (dual-task). Sixteen passes over a 4.88 m walkway in each condition were recorded and step length and duration variability was analysed. Free-living gait was assessed with a waist-worn accelerometer by identifying gait bouts of at least one min duration, and the mean multiscale sample entropy in one mins non-overlapping epochs is reported. Statistical inferences were based on analysis of variance using sex and group as the factors. Results: No difference between those with subjective memory complaints and those without were observed in either laboratory- or free-living gait variability estimates. Both laboratory- and free-living gait variability were higher in those with mild dementia compared to the other groups. Significance: Assuming that subjective memory complaints are on the pathway from cognitively intact to cognitively frail, the findings raise the hypothesis that subjective memory complaints occur earlier in the pathophysiology than measurable changes in laboratory or free living gait. Alternatively the gait variability assessments utilised may have been too insensitive.