LAUSR

OBJECTIVE To identify the occurrence of recurrent falls and the determinants in the acute phase post stroke that are associated with recurrent falls within the first year after stroke onset. DESIGN Prospective follow-up study. SETTING Stroke unit and community. PARTICIPANTS 504 patients with acute stroke. INTERVENTIONS Not applicable. MAIN OUTCOME MEASURES The dependent variable was recurrent falls, defined as two or more falls, within the first year after stroke onset. The independent baseline variables were related to function, activity, participation, personal and environmental factors and comorbidity and were assessed within four days after admission to a stroke unit. Fall data were registered at the stroke unit and self-reported fall data were collected during follow-up using a standardized questionnaire. Determinants of recurrent falls were identified using univariable and multivariable logistic regression analyses. RESULTS Within 12 months after stroke onset, 95 of 348 participants (27%) had experienced recurrent falls. Poor postural control (odds ratio [OR] 5.85, 95% confidence interval [CI] 2.84-12.02, P<.0001), moderate postural control (OR 2.41, 95% CI 1.21-4.80, P=.012) and using a walking aid in the acute phase (OR 2.51, 95% CI 1.45-4.36, P=.0010) are statistically significant determinants that are associated with recurrent falls within the first year after stroke onset. The determinant of using a walking aid appears to be primarily driven by those younger than 80 years. In addition to impaired postural control and using a walking aid, a fall at the stroke unit is a determinant associated with recurrent falls after discharge within 6 months after stroke onset. CONCLUSIONS More than one in four individuals with stroke experienced recurrent falls within the first year after stroke onset. Impaired postural control, using a walking aid in the acute phase and fall during hospitalization are determinants associated with recurrent falls during follow-up. The determinants differ somewhat at different ages.