A taxonomy of cognitive tasks to evaluate cognitive-motor interference on spatiotemoporal gait parameters in older people: a systematic review and meta-analysis

BACKGROUND: Walking in natural environments can be considered a dual-task (DT) scenario that requires increasing cognitive resources with advancing age. Previous reviews concluded that gait speed under DT conditions is equivalent to gait speed as a single task (ST) in the prediction of future falls in older people. However, without a clear taxonomy, these conclusions might be premature. The aim of this review is to use a taxonomy for classifying cognitive tasks of cognitive-motor interference (CMI) paradigms while walking to identify which task domains lead to more pronounced cognitive-motor decrements due to fall risk and concern about falling (CoF) in older people. METHODS: A systematic literature research following PRISMA guidelines was conducted using MEDLINE, Psych-Info and EMBASE. Inclusion criteria were: older people ≥60 years with a previous fall or CoF, use of a DT paradigm to discriminate fallers and non-fallers, straight overground walking, reported gait measurements during ST and DT conditions. A meta-analysis estimated the effect of DT costs for the cognitive task domain and spatiotemporal gait parameters. RESULTS: N = 3737 studies were found within the databases. Nineteen studies were included (n = 14 for meta-analysis). Fallers and people with CoF showed reduced walking speed for ST and DT conditions. Effects of DT were examined for mental tracking tasks. The combined odds ratio (OR [95% confidence interval]) for fallers vs. non-fallers for ST was 3.13 [0.47, 5.80] with moderate heterogeneity (I (2) = 48%). For DT, the OR was 5.17 [2.42, 7.93] with low heterogeneity (I (2) = 37%). Comparing participants with and without CoF, the OR for ST was 12.41 [9.97, 14.84] with high heterogeneity (I (2) = 85%) and OR for mental tracking DT was 10.49 [7.58, 13.40] with moderate heterogeneity (I (2) = 51%). CONCLUSION: CMI was not significantly different between fallers and non-fallers or people with and without CoF; however, our taxonomy revealed a large variety of cognitive conditions and a higher number of studies using mental tracking tasks, which make it impossible to draw firm conclusions. Future studies should use a more standardised and ecologically valid approach when evaluating the validity of DT gait performance in the prediction of falls, CoF or other age-related conditions. TRIAL REGISTRATION: This review was registered at Prospero with the ID: CRD42017068912.