Balance and Gait Characteristics to Predict Accidental Falls in Community-dwelling Stroke Elderly

Background: Accidental falls result in major physical and psychological morbidity in the elderly, especially stroke patients. To get optimal prevention of stroke falls, it is crucial to identify independent risk factors from multivariate analysis of related balance and gait parameters. Aim: To evaluate the effects of balance and gait characteristics in determining whether or not accidental falls may happen. To investigate if the balance and gait parameters affect the time from lesion to when (and if) patients suffer from falls. To compare differences of balance and gait parameters in accidental fall patients between those with severe accidental injuries and those without severe injuries. Methods: The 2-year observational cohort study will enroll 706 elderly stroke patients, who have not had a previous stroke. They will be retrieved from the Inpatient department and followed-up regularly in the Outpatient department when they are discharged. All subjects will be evaluated at 1 month and 6 months after stroke. These evaluations will include: mental, cognition and depression status; lower-limb muscle strength assessed by a dynamometer as well as range of motion of joints assessed by a biaxial electro-goniometry; functional independence measure; balance and gait pattern will be assessed by both clinical observation and objective laboratory examination (a dynamic posturography and computerized gait analysis system); falls and fall-related injuries during the proceeding year. Differences in baseline characteristics are tested by one-way ANOVA, Mann Whitney U or t tests, if appropriate. The investigators will identify the independent risk factors of accidental falls with logistic regression and use the log rank statistic of Kaplan-Meier survival curve to differentiate the risk of accidental falls or fall-related injuries in different groups.