RR-PO-0531

Wednesday 12:30, Palau de Congressos, Exhibition Hall  [Display No. 502]

APPLICATION OF LATENT CURVE MODEL FOR DECLINE IN ACTIVITIES OF DAILY LIVING OF PATIENTS WITH A MORE THAN ONE-YEAR HISTORY OF STROKE.  Harada K, Saitoh K, Tsuda Y, Kagawa K, Nakajima K, Takao Y, Nara I;  Graduate School of Medical Science. Hiroshima University. Hiroshima. Japan. PURPOSE: While epidemiological studies have shown the number of survivors who are independent in activities of daily living (ADL) decreases after the first year post-stroke, there is a lack of knowledge in terms of the progression of decline in the ADL level. The purpose of this study is to test whether the degree of ADL decline late after stroke is influenced by their hypothesized prognositic factors according to a priori empirical findings in former studies.  RELEVANCE: The present study provides understanding of the overall pattern of deterioration in ADL ability for long-term period, and could be used to predict long-term functional outcome.  SUBJECTS: Subjects were outpatients with a more than one-year history of a stroke at one local hospital. Sixty-four outpatients (37 male), who provided full data set throughout this prospective longitudinal survey and also had adequate cognitive function, were included. The average age at baseline survey was 71.7.  METHODS: ADL was measured twice over a 12-month interval by the 5 item Barthel index (Hobart & Thompson, 2001); total score has the high possibility of detecting intra-individual change over time. Dependent variables were intercept factor (the initial level of ADL) and slope factor (the degree of subsequent ADL annual change). Predictor variables were age at baseline survey and Brunnstrom recovery stage of lower limb.  ANALYSES: Latent curve analysis method from repeated-measures data was applied to predict longitudinal linear change in ADL using AMOS. Descriptive analyses were conducted using SPSS.  RESULTS: The decrease in ADL score of the 5 item Barthel index between the two measurements was found to be significant (the Wilcoxon matched-pairs signed-rank test, p < 0.001). Model fitting indices (p > 0.05, the Comparative Fit Index = 0.997) indicated a good fit. Increasing age and persistence of severe motor dysfunction were both significantly associated with a larger degree of annual ADL decline.  CONCLUSION: Advanced age and severe motor deficit can interact in determining the degree of ADL deterioration in patients with a more than one-year history of a stroke. The model contributes to identify long-term survivors after stroke at risk of faster ADL decline.