LAUSR

Background Malnutrition is confirmed to be associated with poor outcomes in stroke patients. The present study aimed to confirm that being at risk of malnutrition assessed by Nutritional Risk Screening Tool 2002 (NRS-2002) and the Controlling Nutritional Status (CONUT) score predicts poor outcomes at 3 months in acute ischemic stroke (AIS) patients. Methods In total, 682 patients with AIS were recruited within 7 days of stroke onset consecutively and 110 were dropped out. They were screened for risk of malnutrition using NRS-2002 and the CONUT score. The primary outcome is the follow-up modified Rankin Scale (mRS) score. Poor outcomes were defined as an (mRS) score ≥ 3 at 3 months post discharge. Results There was a significant difference in the mRS score at 3 months between patients at risk of malnutrition compared to those not at risk assessed by NRS-2002( P  < 0.001) and CONUT ( P  = 0.011). The logistic regression model showed that the risk of malnourishment (according to NRS-2002), low risk of malnourishment (according to CONUT), and the moderate-to-severe risk of malnourishment (according to CONUT) were associated with higher risk of poor outcomes at 3 months ( P  < 0.001, P  = 0.033, and P  = 0.007). The multivariate logistic regression model (adjusted for confounding factors) demonstrated that the risk of malnourishment, according to the NRS-2002, was associated with the increasing risk of poor outcomes at 3 months (odds ratio = 2.31; 95% CI: 1.24–4.30; P  = 0.008). Conclusions The risk of malnutrition assessed by NRS-2002 and CONUT can predict poor outcomes at 3 months in AIS patients. NRS-2002 is superior to CONUT in predicting poor outcomes at 3 months.