Psychometric Properties of Four Step Tests and Effects of Step Training in Reducing Falls on Older Adults with Dementia

Abstract: Objectives: (i) To investigate the test-retest and inter-rater reliability, construct and known-group validity of four independent step tests in older adults with mild to moderate dementia (Phase 1); and (ii) the feasibility and effects of a 12-week step training on the step performance, physical and functional outcomes in this population (Phase 2). Design and subjects: Older adults who have mild to moderate dementia, and able to walk independently without walking aids, with stick or quadripod will be recruited to participate in this study. Participants will perform four step tests, including Four Square Step Test (FSST), Choice Stepping Reaction Time Test (CSRTT), Maximum Step Length Test (MSLT) and Alternate Step Test (AST) on three separate testing occasions conducted by two independent raters at the baseline (Phase 1). The participants will then join either an intervention or control group. Those in the intervention group will receive a 12-week step training, while the control group will receive usual care. The stepping performance using the step tests validated in Phase 1, cognitive and functional outcomes and prospective falls will be compared between the intervention and control groups at 12 and 24 weeks (Phase 2). Interventions: The step training consists of two 30-minute training sessions per week. The participants will receive an individualized, progressive training that is tailored to their stepping ability and cognitive function. Main outcome measures: Stepping performance using the four step tests, 2-minute walk test, 10-meter walk test, 30-second sit to stand test, Berg Balance Scale, Montreal Cognitive Assessment and prospective falls at 12 and 24 weeks. Expected results: (i) The four step tests are reliable in older adults with mild to moderate dementia, and the findings of the step tests are moderately associated with the functional outcomes, and (ii) the participants in the intervention group will have better stepping performance, better functional outcomes and fewer falls compared with the control group at 12 weeks, and the effects will sustain at 24 weeks.