Impact of pre-stroke frailty on outcome three years after acute stroke: The Nor-Coast study
Munthe-Kaas, Ragnhild; Lydersen, Stian; Quinn, Terry J.; Aam, Stina; Pendlebury, Sarah T.; Ihle-Hansen, Hege Beate
Abstract
Introduction: We aimed to explore the predictive value of pre-stroke frailty index (FI) on functional dependency and mortality 3 years after stroke. Methods: Based on the Rockwood 36-item FI score, we calculated the pre-stroke FI from medical conditions recorded at baseline in the multicenter prospective Nor-COAST study 2015–2017. Participants with a FI score and a modified Rankin scale (mRS) 0–6 3 years post-stroke were included in this study. We used logistic regression analysis with unfavorable mRS (over 2 vs. 0–2) at 3 years, or dead within 3 years, as dependent variable, and frailty and pre-stroke mRS, one at a time, and simultaneously, as predictors. The analyses were carried out unadjusted and adjusted for the following variables one at a time: Age, sex, years of education, stroke severity at admission, infections treated with antibiotics and stroke progression. We report odds ratio (OR) per 0.10 increase in FI. Results: At baseline, the 609 included patients had mean age 72.8 (standard deviation [SD] 11.8), 261 (43%) were females, and had a FI mean score of 0.16 (SD 0.12), range 0–0.69. During 3 years, 138 (23%) had died. Both the FI, and pre-stroke mRS, were strong predictors for unfavorable mRS (OR 4.1 and 2.7) and dead within 3 years (OR 2.2 and 1.7). Only adjusting for age affected the result. The OR for pre-stroke mRS decreased relatively more than the OR for FI when entered as predictors simultaneously. Conclusions: FI is a stronger predictor than premorbid mRS for prognostication after stroke.