1 Aging, Mobility, and Cognitive Health Laboratory, University of British Columbia, Vancouver, BC, Canada.
2 Djavad Mowafaghian Centre for Brain Health, University of British Columbia, Vancouver, BC, Canada.
3 Department of Physical Therapy, Faculty of Medicine, University of British Columbia, Vancouver, BC, Canada.
4 Centre for Aging SMART at Vancouver Coastal Health, Vancouver Coastal Health Research Institute, Vancouver, BC, Canada.
5 Faculty of Management, University of British Columbia Okanagan, Kelowna, BC, Canada.
6 Department of Kinesiology, Health, and Applied Physiology, Concordia University, Montreal, QC, Canada.
7 Aging, Mobility, and Cognitive Health Laboratory, University of British Columbia, Vancouver, BC, Canada. teresa.ambrose@ubc.ca.
8 Djavad Mowafaghian Centre for Brain Health, University of British Columbia, Vancouver, BC, Canada. teresa.ambrose@ubc.ca.
9 Centre for Aging SMART at Vancouver Coastal Health, Vancouver Coastal Health Research Institute, Vancouver, BC, Canada. teresa.ambrose@ubc.ca.
10 Department of Physical Therapy, Faculty of Medicine, University of British Columbia, Vancouver, BC, Canada. teresa.ambrose@ubc.ca.
11 Djavad Mowafaghian Centre for Brain Health, c/o Liu-Ambrose Lab, 2215 Wesbrook Mall, Vancouver, BC, V6T 2B5, Canada. teresa.ambrose@ubc.ca.

Background: Common consequences of a stroke include impaired motor and cognitive function, with both being linked to increased falls and frailty. Intra-individual variability (IIV) of cognitive performance, which refers to the within-person trial-to-trial variation in reaction time during cognitive tasks, may be a useful predictor for falls in older adults with chronic stroke.

Objective: To examine whether IIV or "traditional" reaction time (RT) measures of cognitive performance predict falls in older adults with chronic stroke.

Methods: This study is a secondary analysis of a proof-of-concept randomized controlled trial (RCT) among community-dwelling adults with a history of stroke, aged 55 years and older, able to walk 6 m, and without dementia. Residualised intraindividual standard deviation (rISD) was the measure of IIV and mean RT was the "traditional" measure of performance on a computerised Stroop Task. Falls were tracked and adjudicated over six months.

Results: 120 participants with a mean (SD) age of 70 (8) years, and 46 (38%) female participants, experienced a mean of 0.61 (SD = 1.15) falls over 6 months. rISD for the congruent Stroop Task condition predicted falls, such that a one-unit increase was associated with 20.5% increase in fall rate.

Conclusion: The findings suggest that IIV metrics may have the potential in fall risk screening post-stroke. Further research is required to evaluate whether IIV in cognitive performance can be improved via interventions such as cognitive training and physical activity.