|| Checking for direct PDF access through Ovid
To investigate the influence of prestroke physical activity (PA) on acute stroke severity.Data from patients with first stroke were retrieved from registries with a cross-sectional design. The variables were PA, age, sex, smoking, diabetes, hypertension and statin treatment, stroke severity, myocardial infarction, new stroke during hospital stay, and duration of inpatient care at stroke unit. PA was assessed with Saltin-Grimby's 4-level Physical Activity Level Scale, and stroke severity was assessed with the National Institutes of Health Stroke Scale (NIHSS). Logistic regression was used to predict stroke severity, and negative binomial regression was used to compare the level of PA and stroke severity.The study included 925 patients with a mean age of 73.1 years, and 45.2% were women. Patients who reported light or moderate PA levels were more likely to present a mild stroke (NIHSS score 0 to 5) compared with physically inactive patients in a model that also included younger age as a predictor (odds ratio = 2.02 for PA and odds ratio = 0.97 for age). The explanatory value was limited at 6.8%. Prestroke PA was associated with less severe stroke, and both light PA such as walking at least 4 h/wk and moderate PA 2–3 h/wk appear to be beneficial. Physical inactivity was associated with increased stroke severity.This study suggests that PA and younger age could result in a less severe stroke. Both light PA such as walking at least 4 h/wk and moderate PA 2–3 h/wk appear to be beneficial.