Research Objectives To investigate emotional predictors of post-stroke participation Design A prospective cross-sectional study. We performed an independent t-test and Pearson's correlation to identify variables for regression models. Then, we built multiple linear regression models using stepwise selection to determine predictors of post-stroke participation. Setting Community. Participants 71 community-dwelling stroke survivors. Interventions Not applicable. Main Outcome Measures Dependent variables, which are participation measures, were the Stroke Impact Scale (SIS)-Participation/Role Function for model 1, Activity Card Sort (ACS) total percent retained for model 2, and Reintegration to Normal Living (RNL) total score for model 3. Emotional independent variables included SIS Emotion, Visual Analog Mood Scale (VAMS) energetic, and happy subscales, Patient Health Questionnaire-9 (PHQ-9), the State-Trait Anxiety Inventory (STAI)-Trait, and Apathy Evaluation Scale (AES). Results Based on independent t-tests and Pearson's correlations, all six emotional variables were retained in regression models and education, and total NIHSS score for model 1, education for model 2, and total NIHSS score for model 3 (p < 0.05). Regression models showed anxiety (β=-0.227), stroke severity (β=-0.292), and depression (β=-0.301) were significant predictors for the SIS Participation/Role Function (p < 0.05), while happiness (β=0.312) and apathy (β=-0.312) were significant predictors for ACS total percent retained (p < 0.05). Moreover, anxiety (β=-0.231), apathy (β=-0.296), and stroke severity (β=-0.670) predicted RNL total score (p < 0.05). The three models accounted for 23.3∼52.7 percent of the variance in participation. Conclusions Although all three outcome measures were designed to assess participation, they consisted of different activities such as work, social activities etc. and had slightly different methods for ascertaining participation. Our results suggest that different emotional predictors may influence participation, though emotion is an important predictor regardless of participation outcome measure. Author(s) Disclosures None disclosed. To investigate emotional predictors of post-stroke participation A prospective cross-sectional study. We performed an independent t-test and Pearson's correlation to identify variables for regression models. Then, we built multiple linear regression models using stepwise selection to determine predictors of post-stroke participation. Community. 71 community-dwelling stroke survivors. Not applicable. Dependent variables, which are participation measures, were the Stroke Impact Scale (SIS)-Participation/Role Function for model 1, Activity Card Sort (ACS) total percent retained for model 2, and Reintegration to Normal Living (RNL) total score for model 3. Emotional independent variables included SIS Emotion, Visual Analog Mood Scale (VAMS) energetic, and happy subscales, Patient Health Questionnaire-9 (PHQ-9), the State-Trait Anxiety Inventory (STAI)-Trait, and Apathy Evaluation Scale (AES). Based on independent t-tests and Pearson's correlations, all six emotional variables were retained in regression models and education, and total NIHSS score for model 1, education for model 2, and total NIHSS score for model 3 (p < 0.05). Regression models showed anxiety (β=-0.227), stroke severity (β=-0.292), and depression (β=-0.301) were significant predictors for the SIS Participation/Role Function (p < 0.05), while happiness (β=0.312) and apathy (β=-0.312) were significant predictors for ACS total percent retained (p < 0.05). Moreover, anxiety (β=-0.231), apathy (β=-0.296), and stroke severity (β=-0.670) predicted RNL total score (p < 0.05). The three models accounted for 23.3∼52.7 percent of the variance in participation. Although all three outcome measures were designed to assess participation, they consisted of different activities such as work, social activities etc. and had slightly different methods for ascertaining participation. Our results suggest that different emotional predictors may influence participation, though emotion is an important predictor regardless of participation outcome measure.
Read full abstract